Regulatory Roles of Histamine in Pre-Eclampsia

Home , Histamine H1 receptor, Histamine H2 receptor

Placental Genomics: Regulatory Roles of Histamine in Pre-eclampsia

Obed Brew

A commentary submitted in partial fulfilment of the requirements of University of West London for the degree of Doctor of Philosophy

2017

2 Acknowledgements

My sincere thanks to Dr Mark Sullivan and Professor Anthony Woodman for their dedicated and continual support for this work. My sincere thanks also to my family for their patience and endurance.

Acts 20:24

3 Abstract

Aim: Pre-eclampsia (PE) is a multifactorial pregnancy related disorder and a major cause of perinatal mortality. Mothers who develop PE present with clinical symptoms akin to experimentally induced elevated histamine. Maternal blood histamine is elevated, while Diamine Amine Oxidase (DAO) level is diminished in PE. The abnormalities in placental development and functions linked to aetiology of PE have similarities with effects of elevated histamine in other tissues, yet the effects of the elevated histamine on placental function were not directly investigated. Therefore, a series of studies were undertaken and published to increase our knowledge and elucidate our understanding on: (1) the functionality of elevated histamine in the placenta; (2) the causes of the elevated histamine in PE placentae, and (3) of the effects of the elevated histamine on placental gene expression and the implications for PE.

Methods: A series of published high-throughput methodologies have been critically discussed in this commentary. Molecular biotechnology and bioinformatics methodologies such as, ex vivo elevated histamine placental explant model, gene cloning, RT-qPCR, In situ hybridization, microarrays, enzymological analysis, ELISA, Immunohistochemistry, RNA expression array assay analysis, integrated meta-gene analyses, Gene Set Enrichment Analysis, Gene Ontology and biological pathway analyses, Leading Edge Metagene analysis, Systematic Reviews with Meta- analysis and Causal effect analysis have been discussed. The rationale for the appropriate use of these methods to investigate the topological expression of placental histamine receptors, regulation of histamine synthesis and production, the effects of elevated histamine on gene expression, and functional roles of the elevated histamine regulated genes in human placenta is also critically discussed.

Results: The findings show that defective Histamine-DAO-Axis (dHDA) underpins PE, and this defect may be precipitated in early pregnancy, thus predating the onset of the clinical manifestation of PE in mothers who later develop the complication; histamine receptors H1 and H2, and DAO messages are expressed in juxtapositions at the foeto-maternal interface, Histidine Decarboxylase (HDC) HDC activity is elevated in PE placentae; and while elevated histamine up- regulates the proteins for Th1-like cytokines, it also down-regulates DAO message expression after prolonged exposure in the placenta. The findings further show that lipopolysaccharides and pro-inflammatory cytokines including IL-10 and INF- increase histamine production in the placenta; and the histamine has a positive feedback loop regulatory relation with the pro- inflammatory cytokines.

Furthermore, the validation of Elevated Histamine Model (EHM), an in vitro model designed for studying histamine effect in placenta showed that placental micro explants (∼50 mg) in long-term culture (explants that have undergone syncytiotrophoblast regeneration) at the liquid-gas interface in 8% oxygen is an optimum culture condition to study effects of histamine in the placenta. EHM produced RNA with quality akin to time zero pre-culture explants.

The works also led to the identifications of a core set of significant genes that are consistently expressed in both normal (NP) in PE placentae but at varying levels, and a further subset of significant genes expressed consistently only in PE placentae (PE specific genes). Comparison of EHM significant genes with the PE specific genes confirmed the presence of 270 consistently expressed genes that appear to underpin the effects of elevated histamine in the placenta with implications for PE pathogenesis. Further analyses of these EHM genes in PE placentae showed that the natural process by which pre-eclampsia develops is affected by the amount of histamine in the maternal blood and placenta, and the natural process by which pre-eclampsia develops is indirectly affected by histamine via covariate functional groups regulated by specific histamine regulated PE placental genes.

4 Conclusion: Histamine receptor genes are expressed at the foeto-maternal interface; DAO genes are expressed in the placenta; HDC activity levels are increased in PE placentae; there is cross-talk between histamine, DAO and cytokines in the placenta; elevated histamine regulated the expression of specific genes in the placenta and these genes are abnormally expressed in PE placentae; the functions of the histamine regulated genes also identified in PE placentae are involved in tissue morphology and possibly poor placentation, metabolic defects, endothelial dysfunction, inflammation, immunologic response, angiogenic and anti-angiogenic response in PE placentae. Therefore it is reasonable to conclude that the elevated histamine observed in PE would have pathophysiological roles in PE and early detection leading to effective control of maternal blood histamine levels has survival values and it’s thus recommended.

5 Table of Contents

Contents Abstract ...... 4 Table of Figures ...... 9 List of Tables ...... 9 List of Abbreviations ...... 10 Chapter 1 ...... 12 Introduction ...... 12 1.1 Background ...... 12 1.2 Theories of Pre-eclampsia Pathophysiology and the Role of Histamine ...... 14 1.3 How could Histamine be the missing link in Pre-eclampsia Pathogenesis? ...... 16 1.4 Academic Biography ...... 21 1.5 Portfolio of Publications ...... 23 1.6 Themes for Commentary ...... 25 Chapter 2 Theme 1: Histamine links with Pre-eclampsia: Evidence from Systematic Reviews ...... 26 2.1 Introduction ...... 27 Account of originality at the time of publication ...... 27 2.2 Contributions made by the reviews to the understanding of the impact of Maternal Blood Histamine on pre- eclamptic pregnancy outcome...... 28 2.2.1 The Roles of Pre-formed and De Novo Histamine in Pregnancy ...... 28 2.2.2 The Sources of Histamine during Pregnancy ...... 30 2.2.3 Histamine Metabolism in Pre-eclampsia ...... 31 2.2.4 Association of diminished maternal DAO activity and increasing blood histamine with Pre-eclampsia .. 35 2.3 Conclusion ...... 41 Chapter 3 Theme 2: Methodological Considerations- Model for studying ex vivo placental gene expression in response to histamine ...... 42 3.1 Introduction: ...... 43 3.2 Sub-section (1): Contributions made to the knowledge and understanding of Patients, the Placenta and Tissue Culture (#2 - #11) ...... 44 3.2.1 The Patient ...... 44 3.2.2 Justification for using Placental Explants ...... 45 3.2.3 Tissue sampling ...... 46 3.2.4 Effects of Culture Duration, Explant Size and Oxygen on Explant Morphology (Publication #3) ...... 47 3.2.5 Modelling Defective Histamine-DAO-Axis (publications #9 - #11) ...... 50 3.3 Sub-section 2: Justification for evolution of methodological techniques and contributions knowledge: - Transition from Single Gene to Multi-gene Analysis ...... 52 3.4 Sub-section 3: Microarrays and Bioinformatics: Considerations, Rationale and Contributions to Knowledge of Placental Gene Expression ...... 53 3.4.1 Microarray Hybridisation (Publications #4, #9 and #11) ...... 53 3.4.2 Rationale for using Loop microarray design (Publications #4, #9 and #11)...... 54

6 3.4.3 Dealing with the impacts of variances (Publications #4, #9 and #11) ...... 57 3.4.4 Validation of the Microarray Design: Biological Relevance of Pre-culture explant as baseline control for histamine treated cultured explants (Publication #3 & #4) ...... 59 3.4.5 Rationale for Bioinformatics methodologies (Publications #2, #4 and #11) ...... 65 3.5 Conclusion...... 70 Chapter 4 Theme 3: Understanding causes of elevated Histamine in Pregnancy and Pre-eclampsia ...... 72 4.1 Introduction: ...... 73 4.2 HDC activity in PE placentae (Publication #5) ...... 74 4.3 Effects of Cytokines and Endotoxins on Histamine Production in the Placenta (Publication #6)...... 76 4.4 Histamine receptors and DAO mRNA are expressed in human placenta (Publication #7) ...... 79 4.5 Conclusion...... 80 Chapter 5 Theme 4: Effects of histamine on placental genes expression: Implications for pre-eclampsia ...... 81 5.1 Introduction ...... 82 5.2 Histamine has regulatory feedback loop with Placental DAO (Publication #6 & #8)...... 83 5.3 Histamine regulates placental Th-1/Th-2 regulation (Publication #6) ...... 85 5.4 Uncovering the complexities of PE genetic pathways (Publication #2) ...... 89 5.5 Histamine regulates specific genes in pre-eclamptic placenta (Publications #9, 10 & #11) ...... 94 5.5.1 Global Changes in EHM Gene Expression ...... 94 5.5.2 Determination of the plausible effects of histamine in PE Placentae ...... 97 5.5.3 Identification of causal effect of histamine specific genes in PE ...... 103 5.6 Discussions and Conclusion ...... 106 Chapter 6 Final Conclusion ...... 110 6.1 Reflection on contributions made by the publications to the knowledge and Science of Placental Genomics and Pre-eclampsia ...... 110 6.1.1 The placenta expresses histamine receptors, and histamine production and elimination in the PE placentae are impaired ...... 110 6.1.2 Elevated histamine in the placenta has functional roles: it regulates production of Th-1/Th-2 like cytokines in the placenta ...... 112 6.1.3 Optimum culture condition is fundamental for appropriate investigation of histamine effect in human placenta and implications in PE ...... 113 6.1.1 Histamine has complex functions in the placenta: it regulates genetic pathways that have implications in PE placentae ...... 115 6.1.2 Is histamine the missing link in PE pathophysiology? ...... 116 6.2 Reflection on my contributions and my professional development as a research practitioner ...... 119 6.3 Suggestions for future developments ...... 123 6.3.1 Early Maternal Blood Pregnancy Histamine and links with Preeclampsia ...... 123 6.3.2 Histamine Intolerance and Pregnancy Outcome ...... 124 6.3.3 Genetic Variations in histamine pathways and Pregnancy Outcomes ...... 126 6.3.4 Validation of HSPE genes and curation of histamine regulated pathways in PE placentae ...... 126 6.3.5 Histamine effects on CD3 signal transduction in placental T and NK cells ...... 127 Reference ...... 128

7 Appendix ...... 144

Table of Figures

Figure 1-1: Theoretical Concepts for Pre-eclampsia Pathogenesis ...... 15 Figure 1-2: Summary of Systemic Effects of Elevated Histamine ...... 18 Figure 1-3: Conceptual Framework for Histamine in Pre-eclampsia Pathogenesis ...... 20 Figure 2-1: Metabolism of histamine by ascorbic acid ...... 32 Figure 2-2: Mean plot of normal maternal blood histamine levels ...... 36 Figure 2-3: Rodent Oestrous and Gestational Histamine ...... 36 Figure 2-4: Normal and PE Maternal Blood Histamine levels superimposed ...... 39 Figure 2-5: Normal Trimesters and Pre-eclamptic Maternal Blood Histamine ...... 39 Figure 2-6: Normal and PE Maternal Blood Histamine levels Compared ...... 40 Figure 3-1: Microarray Experiment Culture flowchart ...... 54 Figure 3-2: Microarray Experiment Design ...... 56 Figure 3-3: Biological Relevance Network Analysis of Time zero, POC and AOC treated Explants ...... 61 Figure 3-4: Expressed genes in POC and AOC explants ...... 63 Figure 4-1: Cytokines and Endotoxin effect on Placental Histamine Production ...... 77 Figure 4-2: Effects of Cytokines and Endotoxin on Placental histamine production after 24 hours of culture ...... 78 Figure 5-1: Effects of Histamine and Cytokine Synthesis and Production ...... 87 Figure 5-2: Ratio of INF-γ and IL-10 Expression in the Placenta ...... 88 Figure 5-3: Plots of differentially expressed among EHM and Control Samples ...... 96 Figure 5-4: Causal effect identification of Elevated Histamine on Pre-eclampsia ...... 104 Figure 6-1: Histamine fits the missing link in Pre-eclampsia Pathogenesis ...... 118

NB: All figures used or presented here are original, either produced specifically for this commentary or taken from one of the publications submitted here for the award.

List of Tables

Table 2-1: Deamination and Methylation Metabolic Rates in Pregnancy ...... 33 Table 2-2: Deamination to Methylation Ratio...... 34 Table 2-3: Clinical Symptoms of High Blood Histamine and Pre-eclampsia ...... 37 Table 2-4: Normal Histamine and Pregnancy Histamine Levels ...... 38 Table 3-1: Characteristics of Placentae ...... 44 Table 5-1: Differentially Expressed Genes in NP and PE Placentae ...... 90 Table 5-2: Histamine Specific Genes in PE Ontologies ...... 100 Table 5-3: Histamine Regulated Genes Performing 80% of Histamine Mediated Functions in PE ...... 101 Table 5-4: Histamine Mediated Pathways with Leading Edge Genes ...... 102 Table 6-1: Summary of My Contributory ...... 119 Table A-1: Maternal Blood Histamine Levels...... 144 Table A-2: Maternal Blood Histamine Trimester Multiple Comparisons ...... 144 Table A-3: Maternal Urine Histamine after Aminoguanidine Treatment ...... 145

9 List of Abbreviations

The following table is a list of abbreviations commonly used in the thesis. Gene symbols are not included in this list but in the respective gene ID tables. Abbreviation Meaning AA Ascorbic Acid AGE Absolute Gene Expression ANOVA Analysis of variance AOC Atmospheric Oxygen Concentration BP Biological Process BSA Bovine serum albumin CADM2 cell adhesion molecule 2 CC Cellular Component cDNA complementary deoxyribonucleic acid CHE Consistent High Expression CLE Consistent Low Expression C/S Caesarean Section Ct cycle threshold CTAP-III, Connective tissue-activating peptide III DAG Directed Acyclic Graph DAO Diamine Oxidase dH2O Distilled Water dHDA Defective Histamine-DAO-Axis DiAG Diacylglycerol DLK1 delta-like 1 homolog (Drosophila) DMEM Dulbecco’s modified eagle’s medium DMR Deamination to Methylation Ratio DMSO Dimethyl sulphoxide ECL Enhanced chemiluminescence ECM Extracellular matrix EDTA Ethylene diamine tetraacetic acid EHM Elevated Histamine model ESTs Expressed sequence tags EtOH Ethanol EVT Extravillous trophoblasts FDR False discovery rate FFPE Formalin-fixed paraffin-embedded FMI Foeto-maternal Interface GAPDH Glyceraldehyde-3-phosphate dehydrogenase GDA guanine deaminase GEP Gene expression profile GO Gene Ontology GSA Gene Set Analysis GSEA Gene Set Enrichment Analysis H&E Haematoxylin and eosin H1R Histamine Receptor 1

H2O2 Hydrogen peroxide H2R Histamine Receptor 2

10 H3R Histamine Receptor 3 H4R Histamine Receptor 4 HA Histamine hCG Human Chorionic Gonadotrophine hormone HDACs Histone deactylases HDC Histidine Decarboxylase 5(S)HPETE 5(S)-Hydroperoxy-6-trans-8,11,14-cis-eicosatetraenoic acid HG Hyperemesis Gravidarum HMT Histamine n-methyl transferase HrHRF Human Recombinant Histamine-Releasing Factor HRP Horseradish peroxidase HSPE Histamine Specific Genes (expressed) in Pre-eclamptic Placentae (PE) HUVEC Human Vascular Endothelial Cell IHC Immunohistochemistry ImAA Imidazole Acetic Acid ISH In situ hybridisation KEGG Kyoto Encyclopedia of Genes and Genomes LS KS Fisher (LS) and the Kolmogorov-Smirnov (KS) MCP-1, Monocyte chemoattractant protein-1 MCS multiple cloning sites MeHA Methyl-histamine MF Molecular Function MIAA Methyl-imidazole acetic acid MIP-1 a, Macrophage inflammatory protein 1 alpha mRNA Messenger Ribonucleic Acid MSigDB Molecular Signature Database NGF Nerve growth factor NP Normal Placenta NVD Normal Vaginal Delivery NVP Nausea and Vomiting in Pregnancy OMH ratio oxidized histamine/methylated histamine LPS Lipopolysaccharide PAA Phenotype Average Analysis PBS Phosphate-buffered saline PE Pre-eclampsia PLC Phospholipase-C POC Physiological Oxygen Concentration PLP (P5P) Pyridoxal phosphate (pyridoxal-5'-phosphate) PMA Phorbol myristate acetate QC Quality Control RANTES regulated upon activation, normal T cell expressed and secreted RGE Relative Gene Exression RMA Robust Multiarray Average rRNA Ribosomal Ribonucleic Acid RT-PCR Reverse Transcription - Polymersase Chain Reaction SEM Standard Error of the Mean TFT Transcription Factor Targets WebGestalt WEB-based GEne SeT AnaLysis Toolkit

11 Chapter 1

Introduction

1.1 Background Pre-eclampsia (PE) is a major cause of perinatal mortality and it complicates up to

8% of all pregnancies in Western countries (Khan et al. 2006, Duley 2009, WHO

2014). It is one of the top 4 causes of maternal mortality and morbidity worldwide, causing 10 to 15% of maternal deaths (Duley 2009, Say et al. 2014, WHO 2014). PE is characterised by new hypertension (blood pressure of >140/90 mmHg) on two separate readings at least 6 hours apart presenting after 20 weeks' gestation in conjunction with clinically relevant proteinuria (>300mg) per 24 hours and/or with symptoms, and/or biochemical and/or haematological impairment (Tranquilli et al.

2014, NICE 2017).

The symptoms, biochemical and haematological impairment are related to maternal organ dysfunction including renal insufficiency (creatinine >90 µmol/L; 1.02 mg/ dL); liver involvement (elevated transaminases – at least twice upper limit of normal

± right upper quadrant or epigastric abdominal pain); neurological complications

(examples include eclampsia, altered mental status, blindness, stroke, or more commonly hyperreflexia when accompanied by clonus, severe headaches when accompanied by hyperreflexia, and persistent visual scotomata); haematological complications (thrombocytopenia – platelet count below 150,000/dL, DIC, haemolysis) (Tranquilli et al. 2014).

12 PE is a multifactorial disease, and while there is a cautious acceptance of links between familial concordance and maternal polymorphism in the pathogenesis of the disorder (Sutherland et al. 1981, Chesley & Cooper 1986, Arngrimsson et al. 1990,

Cincotta & Brennecke 1998, Esplin et al. 2001, Nilsson et al. 2004, Buurma et al.

2013, Sitras et al. 2015), the placenta is suggested as the primary cause of PE

(Myatt 2002, Matsuo et al. 2007).

The notion denoting the placenta as the main cause of PE is embedded in the fact that the delivery of the placenta is the only known intervention that cures PE (Matsuo et al. 2007). Therefore, the links between PE and placentation is well investigated.

Placentation is basically the process of formation and growth of the placenta and it involves tight regulation of trophoblast proliferation, differentiation and invasion. It is initiated shortly after implantation, when the blastocyst makes contact with the epithelial lining of the uterus (Damsky et al. 1997, Genbacev & Miller 2000). The process enables the trophoblast cells to breach and colonise uterine epithelia, the endometrium and surrounding vasculature to allow nutrient supply to the developing foetus (Pijnenborg 1988, Castellucci et al. 1990, Genbacev & Miller 2000, Reynolds et al. 2010).

The invasion process involves differentiation of the villous cytotrophoblast. Placental villi consisting of inner mesenchymal cells surrounded by a monolayer of mononuclear villous cytotrophoblast stem cells develop (Fisher & Damsky 1993,

Meekins et al. 1994, Lyall 2006, Salamonsen et al. 2009). The cytotrophoblast layer differentiates by either fusing to form the overlying multinucleated syncytiotrophoblast or develops into extravillous trophoblasts (EVT) in anchoring villi

13 (Nelson et al. 1986, Irving et al. 1995, Damsky et al. 1997, James et al. 2005,

Baczyk et al. 2006). The EVT which migrates away from the placenta differentiates further to form invasive extravillous trophoblast (iEVT) (Fisher & Damsky 1993, Zhou et al. 1997, Norwitz et al. 2001).

It seems that a major outcome of invading EVT is the adaptation of the decidua to sustain pregnancy through remodeling of uterine spiral arteries in readiness for increased blood supply required in the second and third trimesters, and sequestration of this blood supply from maternal vasoconstriction (Pijnenborg 1988,

Meekins et al. 1994, Naicker et al. 2003, James et al. 2006, Pringle et al. 2010). The invading EVT thus remodel spiral arterioles by replacing maternal endothelium

(Meekins et al. 1994, Lyall 2003, Pijnenborg et al. 2006a, Pijnenborg et al. 2006b,

Harris et al. 2010, Pringle et al. 2010), and this morphogenesis is suggested to increase the diameter and compliance of the spiral arteries to permit a 12-fold increase in uterine blood flow to meet the demands of the growing foetus in later gestations (Martin 1965, Pringle et al. 2010). It is during the process of placentation that it is believed defective histamine regulation could lead to a cascade of events that trigger PE and some trophoblast related complications of human pregnancy.

1.2 Theories of Pre-eclampsia Pathophysiology and the Role of Histamine

Several theoretical concepts (Figure 1-1A) including abnormal angiogenic and anti- angiogenic factors, poor placentation, defective immunologic response, metabolic disorders, inflammation, and endothelial dysfunction have been proposed to explain the pathogenesis of PE (Mol et al. 2016). The conceptualisation of these theories as proposed to explain the pathophysiology of PE is complex (Redman 2014). Thus to

14 help with the understanding of the pathophysiology of PE, a two-stage model has been advocated (Redman 1991, Sargent et al. 2006, Gammill & Roberts 2007).

Stage 1 speculates that the initiating perturbation of PE is a relative deficient placentation. The deficient placentation when established is suggested to lead to the precipitation of systemic inflammatory response that consequently results in the maternal pathophysiology of PE (Stage 2) (Redman et al. 1999, Roberts & Hubel

2009).

Figure 1-1: Theoretical Concepts for Pre-eclampsia Pathogenesis Figure A: This is a theoretical conceptualisation of the pathophysiological changes associated with pre-eclampsia (PE); Figure B depicts the missing link in the pathophysiology of PE. Figure A is based on Redman 1991, Redman et al. 1999, Sargent et al. 2006, Gammill & Roberts 2007, Roberts & Hubel 2009, Redman 2014, Ilekis et al. 2016, Mol et al. 2016. Figure B is based on my re- interpretation of Figure A.

15 While it is accepted that the pathogenesis of PE is multifactorial and defective placentation plays critical role in the origins of PE (Ilekis et al. 2016), the changes or factors that link the pathophysiological presentations in the placenta to the clinical manifestation of PE have not been fully understood. In other words, it is acknowledged that stage 1 events are not sufficient to cause the stage 2 downstream maternal syndrome (Roberts & Hubel 2009), and that there must be a missing link with plausible functions to interact with both placental and maternal events (genetic, morphological changes, physiological adaptation, behavioural, environmental) to precipitate the clinical manifestation of PE (Figure 1-1B). It was upon the basis of this proposition that it was hypothesised histamine could be the missing link in the pathophysiology of PE, and therefore the ensuing studies commenced.

1.3 How could Histamine be the missing link in Pre-eclampsia Pathogenesis? It was observed that trophoblast related complications of human pregnancy including pre-eclampsia, preterm labour (PL), spontaneous abortion (SA) and hyperemesis gravidarum (HG) present with elevated blood histamine and perilous but varying pregnancy and maternal outcomes (Kapeller-Adler 1941, Kapeller-Adler 1949,

Southren et al. 1963, Beaven et al. 1975, Dubois et al. 1977, Minagawa et al. 1999,

Redman et al. 1999). Histamine has a well-characterised role in the pathogenesis of emesis (Brunton 1996), and experimental elevation of histamine in humans causes nausea and vomiting with a clear relationship between dosage of infused histamine and severity of nausea and vomiting (Kapeller-Adler 1941, Ind et al. 1982). In addition to emesis, intravenous infusion of histamine dose dependently causes rise in heart (pulse) rate, temperature and flushing, or fall in diastolic pressure, and feeling of constriction in the head, headache (minor and severe depending on dose)

16 with or without visual disturbances, cyanosis, inspiratory dyspnoea, convulsions and unconsciousness (Kapeller-Adler 1941, Ind et al. 1982, Kaliner et al. 1982). These symptoms are common in HG and PE, however no further commentary will be provided on the links between histamine and HG. Similarly, pre-term labour presents with elevated histamine (Kirkel' et al. 1983, Caldwell et al. 1988, Riedel et al. 1989,

Donahue et al. 1995, Kramer et al. 1995, Wen et al. 2001). While the pathophysiology of pre-term labour involves inflammatory processes similar to PE, the aetiology of pre-term labour is different from PE (Bytautiene et al. 2004a, Steel et al. 2005, Romero et al. 2014), and therefore no further consideration is also given to it in this commentary. Indeed, these complications of human pregnancy have in common precipitation of histamine-like inflammatory response as underlying pathogenesis (Regan 1992, Sacks et al. 1998, Minagawa et al. 1999, Sacks et al.

1999), nonetheless, due to major differences in the respective aetiology further discussion is herewith focused on histamine and PE.

Histamine is a pleiotropic amine, which appears in tissues of most vertebrates and invertebrates, and the histamine forming capacity of tissues with a net whole blood histamine flow within 0.1M range (plasma equivalent in 2.0nM range) is associated with various physiological functions such as gastric acid secretion, immuno- modulation, cell proliferation and differentiation, tissue growth and wound healing

(Beaven 1982, Falus & Meretey 1992, Kahlson et al. 1960a, Rivera et al. 2000,

Tetlow & Woolley 2003). In humans, the normal blood histamine follows a biorhythmic change (Rehn et al. 1987), which is maintained by a balanced (Section

2.2.2 below) between the activities of the synthesising enzyme histidine decarboxylase (HDC 4.1.1.22), and the metabolising agents including the copper

17 containing diamine amine oxidase (DAO: EC 1.4.3.6), histamine N-methyltransferase

(HNMT: EC 2.1.1.8) and plasma ascorbic acid (Beaven 1982, Clemetson 1980,

Kahlson & Rosengren 1971, Wantke et al. 1998) (section 2.2.3 below).

The elevated blood histamine presents with both clinical symptoms and pathological effects including: oedema, hypertension, proteinuria, nausea & vomiting, headaches, platelet coagulation, endothelial damage, oxidative stress, dysregulation of glucose and lipid metabolism and immune reaction (summarised in Figure 1-2) (Horakova et al. 1977, Beaven 1978, el-Gendi et al. 1988, Shterental' et al. 1991, Thomas et al.

1991, Ebeigbe & Talabi 2014). These histamine effects are akin to maternal clinical manifestation or symptoms of PE.

Figure 1-2: Summary of Systemic Effects of Elevated Histamine Figure shows known effects of elevated histamine in human tissues

Also, randomised clinical trials of histamine augmentation confirmed that elevated histamine prevents inhibition of T cells and natural killer cells by monocyte-derived reactive oxygen metabolites and synergizes with IL-2 to up-regulates Th1 cell

18 responses in patients with melanoma (Asemissen et al. 2005). This report was revelatory, because it not only confirmed that ligand overload triggers dose dependent histamine post-receptor cascade of immunological responses but also supported a previous argument that elevated histamine up-regulates in vivo, thymic and pro-inflammatory cytokines as key pathological pathways (Falus & Meretey

1992, Arad et al. 1996, Panina-Bordignon et al. 1997, van der Pouw Kraan et al.

1998, Darmochwal-Kolarz et al. 1999, Hellstrand et al. 2000, Jutel et al. 2001).

Considering that Natural Killer (NK) and extrathymic T cells are abundant in the human placenta, and numerical changes and functional activation of NK and T cells are more prominent in the blood and urine of patients with PE than in normal pregnant women (Regan 1992, Sacks et al. 1998, Darmochwal-Kolarz et al. 1999,

Minagawa et al. 1999, Redman et al. 1999, Saito et al. 1999); and that the activation of granulocytes, NK cells, and extrathymic T cells are suggested as essential for the maintenance of pregnancy but the over-activation thereof may be responsible for the onset of pregnancy disorders (Wegmann et al. 1993, Minagawa et al. 1999, Saito et al. 1999, Weetman 1999, Saito & Sakai 2003). It was asked whether the elevated histamine in PE could have effects, similar to that observed in peripheral tissues

(Figure 1-2) and thus plausibly function as the missing link in the PE pathophysiology.

Figure 1-3: Conceptual Framework for Histamine in Pre-eclampsia Pathogenesis Figure depicts the theoretical framework linking histamine to PE pathophysiology

To answer this question, it was hypothesised that for histamine to have effects in the placenta, histamine receptors should be expressed at the foeto-maternal interface, and the histamine receptors thus expressed would regulate placental gene expression (Figure 1-3). It was further hypothesised that for histamine to be established as the missing link in the pathophysiology of PE, elevated histamine would regulate specific genes that are also expressed in pre-eclamptic placentae, and the ontologies of these genes would be associated with pathophysiological changes implicated in pathogenesis of pre-eclampsia.

20 1.4 Academic Biography With over 29 years’ experience in healthcare and higher education (HE) including at least 16 years in senior positions in HE, I have made notable contributions to healthcare workforce development through both undergraduate and postgraduate biomedical science, nursing and medical workforce training. I studied Human

Physiology and Pharmacology at the University Hertfordshire and later, Human

Reproductive Biology at the Imperial College London. Clinically, I undertook training in nursing, midwifery and clinical andrology, and worked for over 10 years in both public and private healthcare institutions in London including St Samaritan’s Hospital for women, London, Edgware General Hospital London, St Mary’s Hospital

Paddington, London and the NHS Direct.

Except for a brief period of time working as a visiting researcher at the then Institute of Obstetrics and Gynaecology of the former Royal Postgraduate Medical School in

Queen Charlotte's and Chelsea Hospital, I have over the past 20 years worked sorely in HEI, mainly at the University of West London with a focus on teaching and research. My early career in HE started as basic science tutor for pre-registration nurses, focusing on teaching Physics and Chemistry as applied to Nursing and later as a lecturer in Life and Applied Sciences at Wolfson Institute of Health and Human

Science. In later years, my teaching is mainly focused on precision and personalised medicine, genomics, bioinformatics, service design and improvement in healthcare,

I focus my technical expertise in two predominant areas: in the identification of novel diagnostic biomarkers and therapeutic targets for complications of human pregnancy and in development of high quality pedagogy to provide accessible world class

21 education. In these areas, I have authored other research and scholarly articles including technical reports, research, systematic reviews, and conference proceedings; some of which are presented here for the award of PhD by publication.

22 1.5 Portfolio of Publications The following papers were therefore published from the investigations undertaken to answer the questions above and they form the basis for the award of PhD by

Published work. The publications fall under the discipline of Genomics and

Bioinformatics. The citations are numbered in the order in which they appear in the commentary rather than by year of publication.

1. Brew O, Sullivan MH. (2006) The links between maternal histamine levels and complications of human pregnancy. J Reprod Immunol. Dec;72(1-2):94- 107 2. Brew, O., Sullivan, M.H. & Woodman, A. (2016) Comparison of Normal and Pre-Eclamptic Placental Gene Expression: A Systematic Review with Meta- Analysis. PloS One, 11(8), p.e0161504. 3. Brew, O., Nikolopoulou, E., Hughes, A., Christian, M., Lee, Y., Oduwole, O., Sullivan, M.H.F. & Woodman, A. (2016) Quality of placental RNA: Effects of explant size and culture duration. Placenta, 46, p.45-48. 4. Brew, O. & Sullivan, M.H.F. (2017) Oxygen and tissue culture affect placental gene expression. Placenta, 55, p.13-20. 5. Brew O, Lakasing L, Sullivan M. (2007) Differential Activity of Histidine Decarboxylase in Normal and Pre-eclamptic Placentae. Placenta. May- Jun;28(5-6):585-7. 6. Brew O, Sullivan, M. H., Roller S (2005), "Regulatory loops between cytokines and histamine in the human placenta", Placenta, vol. 26, no. 8-9, p. A.52 7. Brew, OB & Sullivan MH (2001) Localisation of mRNAs for diamine oxidase and histamine receptors H1 and H2, at the foeto-maternal interface of human pregnancy, Inflammation Research, 50(9): 449-452 8. Brew, O. & Sullivan, M.H.F. (2007) Histamine regulates placental diamine oxidase mrna expression - Evidence for a feedback loop decreasing histamine production in pregnancy? Placenta, 28(8 - 9), p.A71. 9. Brew O, Sullivan MHF (2015) Placental Gene Expression in Response to Histamine and Oxygen, Gene Expression Omnibus. Series accession number GSE74446 (http://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE74446) 10. Brew O, Sullivan M. (2017) Elevated Histamine Model: A Protocol for an ex vivo model for in vitro study of histamine effect on placenta. Protocols.io dx.doi.org/10.17504/protocols.io.jigckbw 11. Brew O, Sullivan MHF. Comparison of elevated histamine model and pre- eclamptic placental gene expression, Unpublished

23 Rationale for the inclusion of Publication items #6, #8 and #11

All articles submitted here are currently published as original research except items

#6, #8, and #11. Items #6 and #8 were originally published as conference proceedings respectively at the International Federation of Placental Association

(IFPA) in 2005 Glasgow, UK, and 2007, Kingston, Ontario, Canada. The abstracts from these conference proceedings are published as indicated in the citations. In this document, the findings from these studies are discussed in more detail to provide context to the work assembled herein. Particularly, item #6 supports publications items #5 and #7. Publication item #5 reported on the findings that HDC activity is increased in PE relative to normal placentae. Item #7 on the other hand shows that histamine receptors are expressed in the placenta. Studies reported in item #6 were conducted to test whether elevated histamine has effect in human placenta. The findings are thus discussed in detail in this context to provide unequivocal evidence that elevated histamine regulates the expression of key cytokines implicated in PE.

Similarly, findings from item #8 are discussed below in this context of publications items #9 and #10 to collaboratively demonstrate the evidence that elevated histamine in the placenta not only affect the regulation of key placental proteins but also perturb the expression DAO genes. Item #11 is currently undergoing revision following submission for publication, and the content of this paper is discussed to provide depth and breadth, and contextualisation of the evidence herewith indicated.

24 1.6 Themes for Commentary Commentary is hereby provided with four themes to show the links and coherence between the published works. The themes provide an account for originality of the works at the time of each project or publication, and identify my contributions to the subject of placental genomics, the regulatory roles of histamine and implications in pre-eclampsia. The commentary also provides reappraisal of the originality of the work and reflection about my professional development as a research practitioner and suggestions for future developments. The themes for the commentary are:

1. Histamine links with Pre-eclampsia: Evidence from Systematic Reviews

2. Methodological Considerations- Model for studying ex vivo placental gene expression in response to histamine

3. Understanding causes of elevated Histamine in Pregnancy and Pre- eclampsia

4. Effects of histamine on placental genes expression: Implications for pre- eclampsia

25 Chapter 2 Theme 1: Histamine links with Pre-eclampsia: Evidence from Systematic Reviews

In this chapter:

 The commentary is based on the findings from publication item #1:

Brew O, Sullivan MH. (2006), The links between maternal histamine levels and complications of human pregnancy. J Reprod Immunol. Dec;72(1-2):94- 107

 The results from systematic review of evidence that clarified our understanding of modalities of histamine production and regulation in human pregnancy are discussed.

Highlights:  The placenta is the major source of maternal blood histamine during pregnancy.  Maternal blood histamine level in normal early pregnancy (up to gestational

week 8) is similar to pre-pregnancy levels.  Direct oxidative deamination metabolic pathway by DAO, indirect deamination involving HNMT and DAO and hydroxylation involving ascorbic acid are the active histamine metabolic pathways during pregnancy.

 Histamine oxidative deamination by DAO is the dominant metabolic pathway in pregnant compared to non-pregnant women.  In normal pregnancy maternal plasma DAO activity exponentially increase from gestational week 8, and peaks at mid-trimester (~gestational week 24).

Concurrently, maternal blood histamine levels progressively decline also from gestational week 8 with a mid-trimester (~gestational week 24) nadir, and this forms a normal Histamine-DAO-Axis (HDA).  In PE, there is an apparent failure of the HDA and this is termed defective

Histamine-DAO-Axis (dHDA).  dHDA presents with a failure of maternal plasma DAO activity to exponentially rise from gestation week 8, leading to diminished DAO activity and elevated maternal blood histamine.

 Regression analysis of maternal blood histamine in PE suggests first and second trimesters histamine levels are significantly higher in PE compared to levels in normal pregnancy.  The ratio of histamine oxidative deamination to methylation metabolites concentrations in maternal blood or urine has predictive value for early identification of PE.

26 2.1 Introduction

Account of originality at the time of publication In 1910, Dale and Laidlaw published their seminal work on the effect of histamine on pregnancy outcome (Dale & Laidlaw 1910). This work showed for the first time that elevated histamine in uterine tissues: in this case, subcutaneous injection of histamine into cats in late pregnancy causes strong periodic uterine contractions of the tonic type with no expulsive value but strong enough to cause premature placental separation and intrauterine death of litters (Dale & Laidlaw 1910). Several publications since this seminal work showed associations between increasing maternal blood histamine and specific pregnancy complications such as PE.

Phylogenetically, histamine appears in tissues of most vertebrates and invertebrates, and the histamine forming capacity of tissues seemingly correlates with various physiological functions such as gastric acid secretion, immuno-modulation, cell proliferation and differentiation, tissue growth and wound healing (Kahlson et al.

1960a, Tasaka 1991, Falus & Meretey 1992, Rivera et al. 2000). At the point of initiating the systematic reviews, evidence that histamine was produced during pregnancy was not disputed. However, the sources and significance of the amine in human pregnancy was unclear. There was also no previous systematic review of the evidence to demonstrate coherent links between the varying histamine levels and specific complications of human pregnancy. Therefore, a series of systematic reviews with narrative synthesis and or meta-analysis were conducted. Databases

(Medline, Pubmed, EMBASE) were searched with keywords histamine, diamine oxidase, pregnancy, pre-eclampsia and their variants in Jan 2005 and repeated in

April 2006. A total of 208 articles were reviewed to examine the level and extent of

27 associations between maternal blood histamine and specific pregnancy outcomes to answer the following questions:

1. During pregnancy, does elevated maternal blood histamine originate from the

maternal issues as compared to foetal tissues?

2. In pregnant women, does elevated blood histamine cause clinical symptoms?

3. Does blood histamine levels and metabolism during pregnancy differ between

women with normal pregnancy and those who develop PE?

2.2 Contributions made by the reviews to the understanding of the impact of Maternal Blood Histamine on pre-eclamptic pregnancy outcome

2.2.1 The Roles of Pre-formed and De Novo Histamine in Pregnancy The first set of reviews was conducted to determine the sources and the type, i.e., pre-formed and de novo histamine in human pregnancy. It was observed that the nature of histamine production in both vertebrates and invertebrates indicates two main sources for histamine appearance in tissues: pre-formed histamine from storage cells e.g., mast cells, basophils, and platelets; and nascent histamine or histamine produced de novo by induction of histidine decarboxylase (HDC) activity

(Beaven 1982, Kahlson & Rosengren 1971). In both cases the pyridoxal phosphate dependent HDC enzyme metabolises histidine to produce histamine (Reite 1972,

Beaven 1982), and its activity mostly serves as a rate-limiting step for histamine production (Kahlson et al. 1958a, Nilsson et al. 1959). It was also clear that diamine oxidase (DAO), histamine methyl transferase (HMT), Pyridoxal phosphate (PLP), and ascorbic acid (AA) were established as important metabolising agents for both nascent and pre-formed histamine (Cooper & Schayer 1956, Schayer & Karjala

1956, Nandi et al. 1974, Chatterjee et al. 1975).

Using thematic analysis, the review revealed that the pre-formed histamine functions more as a stereoadaptive-defence agent, where in both vertebrates and invertebrates it is mainly involved with immediate defence response attack as in antigeneic or predatory invasion (Reite 1972, Beaven 1982). In these instances the review further showed that pre-formed histamine commonly occurs in secretory glands and in their secretions, e.g. in many poisonous glands; in gastric mucosa, gastric glands and gastric secretions of most eutherians and ancestral vertebrates of the cyclostomes and cartilaginous fish types, like hagfish and dogfish respectively; and in salivary glands of octopus and squid (Reite 1972). These initial findings prompted a subsequent review to establish the most common type of histamine produced during human pregnancy.

The subsequent results showed that: (1) Pre-formed histamine is produced and stored in mast cells in both male and female reproductive apparatus e.g. human testes, seminal vesicles, prostate glands, uterine tissues, placenta and ovaries

(Wicksell 1949, Purcell & Hanahoe 1991, Purcell 1992, Szelag et al. 2002a, Szelag et al. 2002b); (2) That the rate of turnover of pre-formed histamine in maternal uterus and the placenta is quite slow and therefore suggests that pre-formed histamine may have minor physiological roles in early pregnancy or during human pregnancy

(Beaven 1982); (3) That de novo synthesised histamine characterised by HDC activity is prominent in pregnancy and is localised in cleaving embryo, fallopian tube and at implantation sites (Blanco et al. 1970, Bronson & Wallach 1977, Shrivastav et al. 1988, Cocchiara et al. 1987, Cocchiara et al. 1992, Menezo et al. 1993,

Cocchiara et al. 1996, Sawin et al. 1997, Jankovic et al. 1998); (4) While the pre-

29 formed histamine in uterine and endometrial tissues has been associated with muscle contraction and immuno-protection of uterine or endometrial milieu

(Bytautiene et al. 2003, Bytautiene et al. 2004b, Purcell & Hanahoe 1991, Thomson et al. 2000, Wicksell 1949, King et al. 2003), de novo histamine synthesised by endometrial epithelia is shown in animal models and human cell lines to initiate implantation, functions in anabolic processes like regenerative growth, cell differentiation and proliferation, imuno-modulation, placentation, and foetal development (Kahlson et al. 1958a, Kahlson et al. 1958b, Kahlson et al. 1960b,

Connolly et al. 1962, Johnson & Dey 1980, Dey & Hubbard 1981, Hine et al. 1985,

Cocchiara et al. 1987, Rivera et al. 2000, Zhao et al. 2000, Dey et al. 2004, Liu et al.

2004), and hence could have similar roles in human pregnancy.

2.2.2 The Sources of Histamine during Pregnancy It became clear that histamine thus observed in human pregnancy could originate from the uterus, foetal membranes and or from the placenta. Therefore, a further review question was conducted to clarify the source of increasing histamine in human pregnancy. For this review, attention was focused on human studies which have investigated in vivo metabolism of 14C-histamine, labelled at the 2nd position of the imidazole ring in women including during pregnancy.

Narrative synthesis was used to summarise the findings from the included studies

(Schayer 1952b, Cooper & Schayer 1956, Nilsson et al. 1959, Schayer 1959,

Lindberg 1963b, Lindberg 1963a, Lindberg et al. 1963a, Lindberg et al. 1963b,

Lindberg & Tornqvist 1966). The synthesis showed that (1) when labelled histamine was injected directly into maternal cubital vein, less than 5% of the histamine was

30 detected in umbilical vein; (2) when labelled histamine was injected into umbilical artery, less than 0.3% of the histamine was detected in maternal urine; (3) when the labelled histamine was injected into the uterus of pregnant and non-pregnant women, 49% as compared to 36% of histamine was metabolised during uterine passage into maternal blood respectively in pregnant and non-pregnant women; and

(4) that M-methyl imidazole acetic acid (MIAA) resulting from methylation of the nitrogen on the imdazole ring remote from the amine side chain was secreted into maternal circulation from the placenta. These findings therefore suggested that the placenta, and not the uterus or the foetus is the major source of increasing histamine in maternal blood during pregnancy, and thus focused our attention on the placenta as the potential medium where histamine could exert effect on pregnancy outcome.

2.2.3 Histamine Metabolism in Pre-eclampsia After establishing that the placenta is the main source of elevated blood histamine during human pregnancy, a critical question about what causes the elevated histamine in PE pregnancy however remained unclear. Therefore, a further review was undertaken to determine if there was any association between histamine metabolism and pre-eclamptic pregnancy outcome.

Generally, histamine is metabolised via 2 major pathways: methylation and deamination pathways. In the methylation pathway histamine is metabolised via methylation at the imidazole nitrogen furthest from the ethylamine side chain (termed tele-N or Nt) by the enzyme histamine-N-methyltransferase (HNMT) to N-Methyl- histamine (NMH), and monoamine oxidase (MAO) then metabolises the NMH to M- methyl imidazole acetic acid (MIAA) (Schayer 1952b, Cooper & Schayer 1956,

31 Schayer 1956, Schayer & Karjala 1956, Rothschild & Schayer 1958, Schayer &

Reilly 1975). In the deamination pathway, DAO metabolises histamine to imidazole acetic acid (ImAA) (Schayer 1952b).

Using narrative synthesis, the review showed that the methylation and deamination of histamine pathways accounts for about 80% of histamine metabolism during human pregnancy, and hydroxylation of histamine into hydantoin propionic acid by ascorbic acid (Figure 2-1) could account for between 10 -20% (Ahlmark 1944,

Wicksell 1949, Kapeller-Adler 1952a, Kapeller-Adler 1952b, Cooper & Schayer 1956,

Schayer & Karjala 1956, Lindberg 1963a, Bjuro et al. 1964, Kapeller-Adler 1965,

Nandi et al. 1974, Chatterjee et al. 1975).

Figure 2-1: Metabolism of histamine by ascorbic acid Figure 2-1 was written in 2006 based on the systematic review of metabolic effect of ascorbic acid on histamine.

The results also showed that the relative metabolism differed between maternal and non-pregnant women. Table 2-1 summarises the relative concentrations of histamine metabolites from the deamination and methylation pathways. The data (Table 2-1) shows an approximate 33% increase in histamine metabolism via direct deamination by DAO and 58% decrease in histamine metabolism via HNMT methylation during pregnancy compared to non-pregnant women, thus confirming that DAO is the dominant histamine metabolic pathway during pregnancy.

Table 2-1: Deamination and Methylation Metabolic Rates in Pregnancy

Histamine Metabolites Non-pregnant Pregnant women Average Change in women levels levels Pregnancy

MIAA 32 – 59% 20 – 41% -57.69%

ImAA 25 – 37% 26 – 53%, 32.7%

(Schayer 1952b, Nilsson et al. 1959, Lindberg 1963a); MIAA = M-methyl imidazole acetic acid; ImAA = imidazole acetic acid.

Somehow, the confirmation that DAO is the major histamine metabolising enzyme during human pregnancy was pedestrian. Nonetheless, a regression analysis revealed that the deamination to methylation ratio (DMR) of histamine metabolites in maternal blood has predictive value for pregnancy outcome, and this was novel

(Table 2-2). Using a comparative analysis of the metabolites from pregnant and non- pregnant women, it was observed that the ratio of deaminated histamine metabolites

(imidazole acetic acid) and the methylated metabolites (n-methyl-histamine, and methyl-imidazole acetic acid) were 4 and 2 folds respectively in pregnant and non- pregnant women arterial or venous blood (Table 2-2), and this was highly informative

33 of maternal blood capacity to control histamine within range associated with good pregnancy outcome prognosis.

Table 2-2: Deamination to Methylation Ratio Relative abundance of Histamine Metabolites in Maternal Blood and Urine from Diamine Oxidase and Histamine N- Methyltransferase Activities

Sampling and radio-labelled histamine metabolite analysis count (cpm/ml) Radio-labelled histamine metabolite Brachial Artery Cubital Vein Uterine Vein Umbilical Urine Vein

From 4 pregnant women Values in blood sample taken from 6 pregnant women (a) (b) Ratio of ImAA ÷ 2.21 1.98 2.28 1.90 1.31 (NMH + MIAA) From 3 non-pregnant Values in blood samples from 4 non-pregnant women (a) women (b)

Ratio of ImAA ÷ 0.51 0.71 0.34 - 0.75 (NMH + MIAA) cpm = Counts per minutes; N-Methyl-histamine = NMH, Methyl-imidazole acetic acid = MIAA, Imidazole Acetic Acid = ImAA, (a) from (Lindberg 1963a), (b) from (Nilsson, LINDELL, Schayer, & Westling 1959)

The analysis further showed that whilst the measurement of individual histamine metabolites in urine was erratic and non-predictive of pregnancy outcomes, the assessment of the ratio of the metabolites in urine or blood could provide a measure or competence of maternal blood to control histamine levels during pregnancy, and thus could predict plausible variations associated with PE as a plausible non- invasive model. Further work using prospective longitudinal studies however, is needed to test this emerging hypothesis.

In effect, these outputs suggested that there are three instead of two major histamine metabolic pathways during human pregnancy: (a) Direct DAO oxidative deamination metabolic pathway (b) indirect deamination pathway involving HNMT and DAO and

34 (c) hydroxylation pathway involving ascorbic acid. The work also suggested that a further review to examine the variations between maternal histamine metabolism in normal and complicated pregnancies could provide additional information to contribute to explaining the association between maternal blood histamine levels and

PE. Therefore a further review was conducted to examine the mechanisms of histamine regulation during normal and complicated human pregnancies.

2.2.4 Association of diminished maternal DAO activity and increasing blood histamine with Pre-eclampsia A further review using comparative analysis was conducted to examine the characteristics and profiles of maternal blood histamine and plasma DAO activity during normal and PE pregnancy. Twelve studies met the inclusion criteria for this analysis (Connolly et al. 1962, Southren et al. 1966a, Tryding & Willert 1968, Achari et al. 1971, Vanina et al. 1971, Beaven et al. 1975, Dubois et al. 1977, Sharma et al.

1981, Morgan & Hytten 1984, Sharma et al. 1984, Hine et al. 1985, Caldwell et al.

1988). The data synthesis was separated into animal and human profiles. The review showed that during normal pregnancy, maternal blood histamine levels gradually decrease below values found in healthy non-pregnant women (Figure 2-2), and this characteristic decline of gestational blood histamine was conserved in rodent models (Figure 2-3). In human normal pregnancy, the progressive decrease in maternal blood histamine (Figure 2-2) reaches a nadir at mid-gestation (weeks 20 and 24) and this was associated with a concurrent exponential rise in maternal plasma DAO activity from gestational week 8 onwards in normal pregnancies with a peak also at mid-gestational (weeks 20 and 24) (Southren et al. 1966a, Tryding &

Willert 1968, Achari et al. 1971, Beaven et al. 1975, Dubois et al. 1977).

20 Histamine Levels Histamine Levels (ng/ml) 10

0 0 5 10 15 20 25 30 35 40 45 Gestational Weeks

Normal Blood HA

Figure 2-2: Mean plot of normal maternal blood histamine levels (Southren et al. 1966a, Tryding & Willert 1968, Achari et al. 1971, Beaven et al. 1975, Dubois et al. 1977, Sharma et al. 1981, Sharma et al. 1984, Caldwell et al. 1988)

Figure 2-3: Rodent Oestrous and Gestational Histamine Figure shows the profile of histamine production in rodent reproductive tissues across the oestrous cycle (eDay) and gestation (gDay). Data is mean plot of histamine levels pooled from rodent studies (Connolly et al. 1962, Hine et al. 1985).

36 The significance of the low histamine levels in mid-gestation is unclear; however the coincidence with DAO peak activity and the period of uterine tranquillity (Morgan &

Hytten 1984) appears to support the hypothesis that tight control of maternal blood histamine is required for normal pregnancy outcome. Therefore, the review further examined the links between hyper-histaminemia occurring in specific gestational complications including preeclampsia, spontaneous abortion, preterm labour and hyperemesis gravidarum. In the case of PE, it was revealed that the complication presents with clinical symptoms similar to those observed in experimentally induced elevation of blood histamine in humans (Table 2-3).

Table 2-3: Clinical Symptoms of High Blood Histamine and Pre-eclampsia Symptoms of Experimentally Induced High Common Symptoms of PE Blood Histamine in normal healthy human volunteers Dizziness Severe headaches Light-headedness Blurred vision Fainting Seeing flashing lights Tachycardia Severe heartburn Pyrexia Upper abdominal pain just below the ribs Palpitation Nausea Continuous headache Vomiting Severe headache Excessive weight gain Nervousness Decreased urine output (fluid retention) Convulsions (seizures) Feeling very unwell Difficulty breathing Oedema Flushing or redness of face Proteinuria Blue discoloration of face Light sensitivity Blurred vision Decreased levels of platelets in your blood Chest discomfort or pain (thrombocytopenia) Hypotension Impaired liver function Hypertension Shortness of breath, Diarrhoea Nausea Vomiting Metallic taste Oedema Abdominal or stomach spasm or cramps Proteinuria (Ind et al. 1982, Kaliner et al. 1982, Ind et al. 1983, Pollock et al. 1991, Thomas et al. 1991, Schmetterer et al. 1997, Akerman et al. 2002, Lassen et al. 2003, NICE 2008, NICE 2011)

37 Interestingly, the review also showed that the average maternal blood histamine levels in early normal gestation is similar to levels found in non-pregnant women

(Table 2-4), but the levels in normal pregnancy decrease from gestational week 8 onwards to values below that found in healthy non-pregnant women (Figure 2-4 –

Figure 2-6). In contrast, a logarithm regression of 3rd trimester maternal blood histamine level in PE suggests a significantly higher early maternal blood histamine levels compared to normal pregnancy for matched gestational ages (Figure 2-4). The average difference between NP and PE 3rd trimester blood histamine was 18.42%.

Table A-1 & Table A-2 (appendix 1) provide further information on the differences and significance level between normal pregnancy and PE maternal blood histamine.

Similar level of magnitude of change in histamine levels was observed in maternal urine histamine concentration after reanalysis of experimental inhibition of DAO activity with aminoguanidine data (Table A-3; Appendix 1). In this instance, the percentage rise in trimester 1 maternal urine histamine after inhibition of DAO with aminoguanidine was 18.73%.

Table 2-4: Normal Histamine and Pregnancy Histamine Levels Histamine levels in Pregnancy and PE Condition Histamine levels Range Histamine levels (ng/ml) Reference (ng/ml) (Mean ± SEM) Normal Blood 25 – 65.0 56.7 ± 1.85 [1,2] Normal Pregnancy: Trimester 1 58 – 70.0 64.0 ± 6.0 [3] Trimester 2 30 – 50.0 43.3 ± 6.7 [3,4] Trimester 3 35 – 74.3 52.9 ± 1.4 [3,5-11] Pre-eclampsia: Trimester 3 49 – 94.3 69.7 ± 1.9 [8,10]

[1]: (Pfeiffer et al. 1970); [2]:(Lorenz et al. 1972); [3]:(Dubois et al. 1977); [4]: (Clemetson & Cafaro 1981); [5]: Effkemann & Werle 1940, [6]: Kapeller-Adler 1949; [7]: (Gunther & Glick 1967); [8]: (Achari et al. 1971); [9]: (Sharma et al. 1981), [10]: (Sharma et al. 1984); [11]: (Sharma et al. 1985).

120

100

20 Histamine Levels Histamine Levels (ng/ml) 0 0 5 10 15 20 25 30 35 40 45 Gestational Weeks

Normal Blood HA (ng/ml) PE Blood HA (ng/ml) Predicted PE Blood HA (ng/ml)

Figure 2-4: Normal and PE Maternal Blood Histamine levels superimposed Figure is a plot of grand mean of blood histamine levels to depict the pattern of maternal blood histamine in normal (NP) and pre-eclampsia (PE) pregnancies. Black line shows predicted levels of early pregnancy maternal blood histamine in mother who later developed PE. Average rise of early pregnancy blood histamine in PE is projected to be 18.42% above normal. The pattern of gestational dependent variation in blood histamine is similar in both normal and PE pregnancies. (Southren et al. 1966a, Tryding & Willert 1968, Achari et al. 1971, Beaven et al. 1975, Dubois et al. 1977, Sharma et al. 1981, Sharma et al. 1984, Caldwell et al. 1988)

Figure 2-5: Normal Trimesters and Pre-eclamptic Maternal Blood Histamine NP = Normal Pregnancy, PE = Pre-eclampsia

Figure 2-6: Normal and PE Maternal Blood Histamine levels Compared Provide pooled analysis blood histamine means based on: (Achari et al. 1971, Dubois et al. 1977, Sharma et al. 1981, Sharma et al. 1984, Caldwell et al. 1988)

These findings showed that although the placenta is a major source of histamine, it also releases active diamine amine oxidase into the maternal circulation, which metabolises histamine (Purcell & Hanahoe 1991). The placental derived DAO once deported into maternal blood breaks down the histamine released from the placenta in normal pregnancy, leading to physiological levels of histamine in normal pregnancy. Clinically, the activity of the deported placental DAO in maternal blood increases a 1000 fold by mid-gestation in normal pregnancy to form a normal histamine-DAO-axis (Southren et al. 1966b, Beaven et al. 1975, Dubois et al. 1977).

In PE, the DAO activity from the placenta is decreased thus, forming a defective histamine-DAO-axis (dHDA). This results in less histamine metabolism in the maternal circulation and higher systemic histamine levels in PE (Achari et al. 1971,

Vanina et al. 1971, Beaven et al. 1975). Thus, normal Histamine-DAO-Axis (HDA) in pregnancy was defined as the concurrent exponential rise of maternal plasma DAO activity levels from gestational week 8 onwards with a peak at mid-trimester

40 (~gestational week 24), and progressive decline of maternal blood histamine levels below pre-pregnancy levels from gestational week 8 onwards with a mid-trimester

(~gestational week 24) nadir. Defective Histamine-DAO-Axis (dHDA) in pregnancy on the other hand was defined as the failure of maternal plasma DAO activity to exponentially rise from gestation week 8, leading to diminished DAO activity and elevated maternal blood histamine akin to pre-pregnancy levels.

2.3 Conclusion The series of reviews provided comprehensive knowledge and in-depth understanding of the links between histamine and PE. In summary, the studies broadened our understanding on the sources of elevated histamine in pregnancy by revealing that the placenta and not the uterus or the foetus is the major source of increasing histamine in maternal blood during pregnancy. The evidence reviewed was conclusive that methylation and deamination pathways accounts for about 80% of histamine metabolism during pregnancy, and hydroxylation of histamine into hydantoin propionic acid by ascorbic acid accounts for between 10 -20%. The work also affirmed that relative metabolic pathways differed between maternal and non- pregnant women, and DAO is the dominant histamine metabolic pathway during pregnancy. The work further showed that maternal blood histamine levels in early gestation are similar to levels found in non-pregnant women, and the levels in normal pregnancy decrease below values found in healthy non-pregnant women from around gestational week 8 onwards. However, in specific complication of pregnancy such as PE, DAO activity levels diminish after gestational week 8 and histamine levels remain elevated at levels similar to pre-gestational levels, thus forming a defective histamine-DAO-axis (dHDA).

41 Chapter 3 Theme 2: Methodological Considerations- Model for studying ex vivo placental gene expression in response to histamine

In this chapter:

.  The commentary is based on methods used to investigate the works that underpin publication items # 2 – 11.  The commentary discusses the rationale and challenges for using specific

methodologies rather than a detailed discussion or report of the methods in general.

Highlights:  Placental explant size (micro or macro) and duration of culture affect explant RNA quality.  Micro explants (∼50 mg) in optimum long-term culture (have undergone syncytiotrophoblast regeneration) at 8% oxygen produced RNA with quality akin to pre- culture explants. Quality of RNA from explants in early cultures (48 hours), presumably undergoing  syncytiotrophoblast degeneration is compromised.  Macro explants (∼200 mg) after culture produced poor quality RNA and should be avoided.  Gene expression patterns differ between pre-culture and cultured explants

 Atmospheric Oxygen Concentration up-regulates HIF1A transcription target gene set

 Atmospheric Oxygen Concentration regulated genes that favour apoptosis and

inflammation

 Tissue culture up-regulates apoptosis and response to stress genes in placenta

 Tissue culture per se induces expression of genes with preponderance towards

apoptosis, inflammation and tissue response to stress irrespective of culture oxygen

concentration.

 Explant culture at Atmospheric Oxygen Concentration exacerbates tissue expression of stress and programmed cell death genes in placenta.  Oxygen and tissue culture generally precipitate down-regulation of placental genes.

42 3.1 Introduction: It is widely accepted that prudent evaluation and informed selection of choiced research methods are the most appropriate and cost effective approach to answer a research question. The selection of methods and the implementation thereof, are however influenced not only by the research question but also by practical considerations, such as the availability of resources, including the type of data available and the knowledge and skills of the persons undertaking the research, access to funds and time constraints. This chapter likewise, discusses key factors that impacted on decisions, selection, applications and implementation of methods used to investigate the genomic impact of histamine in human placenta. The discussions in this chapter are therefore based on rationales for: (1) Patients and the

Placental choices; (2) Transition from Single Gene Analysis to Multi-gene Analysis;

(3) Microarrays: Study Design Considerations and Rationale, and (4) Bioinformatics analyses. This chapter is therefore organised into three sub-sections as follows:

 In sub-section 1, the commentary focuses on addressing specific questions as: What was the need to be selective about patients? Why use placental explants instead of cell lines? What was the need for sampling tissues? Was the use of short term explant culture appropriate model for ex vivo investigation of histamine effect on the placenta? Blocking of DAO with aminoguanidine disulphate: what was the need?

 In sub-section 2, the focus shifts onto addressing the impact of Technological advancement on the research agenda. Specifically, the section addresses questions related to the need to reshape the research approach in line with evolution of technology from single gene expression analysis to multiplex, and then microarray.

 In sub-section 3, issues related to explant viability after culture and selection of an appropriate baseline control are discussed to show the respective contribution to the knowledge and understanding of placental explant culture.

43 3.2 Sub-section (1): Contributions made to the knowledge and understanding of Patients, the Placenta and Tissue Culture (#2 - #11)

3.2.1 The Patient Human pregnancy has multiple outcomes including spontaneous abortion, PE, and pre-term labour. Therefore, defining the patients involved in the investigations was important to embed the findings in the context of specific complications of human pregnancy. Placental samples were taken from women whose pregnancy outcomes were classified as a first trimester social abortion, normal term, or PE (Table 3-1).

Clinically diagnosed PE was defined in terms of WHO criteria (updated in 2014) as new hypertension (raised diastolic pressure to >90 mmHg on two occasions, 1–4 hours apart) presenting after 20 weeks with proteinuria (>30mg/dL or +1 or more on a single sample).

Table 3-1: Characteristics of Placentae

Clinical Condition Gestation Number of Delivery Foetal birth of Placenta (weeks) placenta Method weight

Normal 8 – 10 3 SA NA

Normal 38 – 40 40 C/S ≥2.5kg ≤ 4.0kg

Pre-eclampsia 38 – 39 10 C/S ≥2.5kg ≤ 4.0kg C/S = Caesarean Section; SA = Social Abortion

Thus, all placentae from normal pregnancies had no history of gestational hypertension or pre-existing hypertension, proteinuria, infection, glucosuria and other pregnancy related complications such as intra-uterine growth restriction.

44 3.2.2 Justification for using Placental Explants The placenta is implicated in specific complications of human pregnancy such as PE

(Myatt 2002, Matsuo et al. 2007). The mature human placenta is a discoid organ 20-

25 cm in diameter, 3 cm thick and weighing 400-600g. Internally, it consists of a foetal villous tree bathed directly by maternal blood (Castellucci & Kaufmann 1982).

The cells making up the internal of the mature placenta consists of several cell types including cytotrophoblast, syncytiotrophoblast, stromal, endothelial cells, pericytes, smooth muscle cells, erythrocytes and macrophages (Hofbauer cells) (Castellucci &

Kaufmann 1982).

Various in vitro models including: choriocarcinoma cell lines such as BeWo, Jeg-3 and Jar cells; isolated primary trophoblast cells, placental mesenchymal stromal cell lines, and villous parenchyma (placental) explants have been used to investigate placental physiology, pharmacology and other functions (Miller et al. 2005, Bode et al. 2006, Orendi et al. 2011, Daoud et al. 2016, Qin et al. 2016). Of all the models however, placental explant is reported to closely mimic in vivo conditions to support ex vivo study of placental functions including cellular uptake, production and release of secretory components, cell interactions, proliferation, growth and differentiation, gene delivery, pharmacology, toxicology, and disease processes (Miller et al. 2005,

Orendi et al. 2011).

Of particular importance to histamine effects, caesarean delivered explants were the first choice of tissue for investigating the roles of histamine in the placenta. This was principally important because, it is observed that placentae with uterine contractile activity produce varying levels of histamine: less in decidua from maternal side, and

45 high levels in tissues from foetal side (Szukiewicz et al. 1995). It is also established that tissue damage and stress destabilise histamine production (Schayer 1952a,

Norn 1968, Neugebauer & Lorenz 1982, Sooranna et al. 2004, Mohan et al. 2007): thereby, making the study of histamine in placentae that have undergone uterine contraction less suitable for mimicking in vivo conditions. Furthermore, the cross-talk between the trophectoderm derived (trophoblasts) and mesenchymal derived cells in the placenta (Lewis et al. 1996) is difficult to mimic or practically absent in cell line models. Also, the cell line models generically show increased baseline expression of pro-inflammatory receptors such as Toll-like receptors and cytokines (Kauma et al.

1992, Lonsdale et al. 1996, Gierman et al. 2015). Thus, rendering cell lines as less attractive model in favour of placental villous tissue explants to investigate the genomic response of the placenta to histamine.

3.2.3 Tissue sampling Fresh placental samples were collected from singleton pregnancies. Placentae were collected from the Imperial College NHS Trust following informed consent from mothers. Ethics permission was granted by the Hammersmith and Queen Charlotte’s

& Chelsea Hospitals Research Ethics Committee. Samples were taken from non- smoking mothers without gestational diabetes or other complications (except PE), who were delivered by elective Caesarean section. Each placenta was visually inspected for signs of excessive tears, necrosis, and infarction, and samples taken from healthy looking areas about 5 cm away from the placental cord with sterile sharp scissors. Approximately 2 cm3 of the placental tissue were randomly cut with sterile scissors from three different sampling sites. Each sample was carefully excised to include an intact chorionic plate, villous parenchyma, basal plate and

46 decidua. All samples were collected from the maternity operating theatre, washed in phosphate-buffered saline (PBS) solution (pH 7.4) containing 10% penicillin, streptomycin and L-glutamine (Sigma) to remove excess blood and blood clots and transported in 150 ml sterile pots containing warm PBS with 10% penicillin, streptomycin and L-glutamine immediately to the laboratory. The placental explants were then aseptically dissected from the villous parenchyma of each sample, fragmented with scissors and incubated within 30 minutes of delivery.

3.2.4 Effects of Culture Duration, Explant Size and Oxygen on Explant Morphology (Publication #3) The use of placental explants as model to investigate placental physiology is complicated by intrinsic variances related to the viability of the tissues in organ culture, duration of the organ culture, oxygen concentration of the culture media and the size of the explants (Miller et al. 2005). Concerning placental culture oxygen, different concentrations (atmospheric oxygen concentration (AOC) and Physiological oxygen concentrations (POC)) have generally been used during culture of term placentae. It has been argued that POC (8%) reflects in vivo physiology, and provides optimal culture conditions for term placental villous explants than AOC

(20%) (Damsky et al. 1993, Genbacev et al. 1996, Caniggia et al. 2000, Huppertz et al. 2003, Miller et al. 2005, Burton et al. 2006).

Some studies have also reported increased deterioration of placental explant viability within 24 hours of culture (Palmer et al. 1997, Sooranna et al. 1999) while others have demonstrated its viability for more than 7 days in organ culture (Tao & Hertig

1965, Taylor & Hancock 1973, Faye et al. 2005). Moreover, placental explant

47 syncytiotrophoblast (STB) cell population are completely regenerated in the course of days in organ culture (Taylor & Hancock 1973, Palmer et al. 1997, Siman et al.

2001). Therefore, the establishment of an appropriate duration and condition for explant culture was crucial, as was determining the baseline control samples for reasons related to the plausible impact of syncytial degeneration on explant RNA quality. Thus, it was aimed to determine whether syncytial degeneration and regeneration could affect the interpretation of results, and also to establish an appropriate explant size, duration of culture and optimum oxygen concentration for the study of histamine effect on human placenta.

In order to evaluate the impact of placental explant size, culture method, culture oxygen concentration and duration of culture on explant viability and quality of RNA placental micro and macro explant in different culture settings were compared.

Placental macro explants (approximately 200 mg wet weight) and micro explants

(<50 mg wet weight) were cultured to time 0h, 24h, 48h, 72h, 96h, 120h and 144h

(publication #3). Microcapillary electrophoretic RNA separation and analysis of automatically generated signal measurements was performed on RNA extracted from placental explants incubated at Physiological Oxygen Concentration (8%) or

AOC in standard or optimum culture (publication #3). Explant viability was indirectly assessed with immunohistochemistry for the STB integrity.

Previous studies have established very clearly that placental tissue, particularly placental explants (Taylor & Hancock 1973, Watson et al. 1995, Tao & Hertig 1965) show loss of the STB layer during initial culture (2-4 days), followed by regeneration of the syncytium over longer culture periods (5 – 11 days) (Palmer et al. 1997, Siman et al. 2001). Therefore, STB integrity was measured by the extent of degeneration

48 and regeneration of the STB layer. STB degeneration was defined as the presence of small, dense and rounded nuclei, irregular syncytial layer with disintegration of the integrity of the apical membrane; detachment and sloughing off of the original syncytial layer in culture (Tao & Hertig 1965, Palmer et al. 1997). STB regeneration was on the other hand, defined according Siman et al (2001) as the progressive replacement of the original STB by a newly formed layer, evidenced by gradual thickening of the new STB as the original syncytium is sloughed off and lost.

The findings (publication #3) showed that integrity of RNA in the macro samples irrespective of oxygen concentration and micro explants cultured in AOC was poor throughout the culture period, whereas in the micro explants cultured in optimal media in POC, the RNA quality was transiently poor, and then improved by day 6 of culture. Thus, suggesting that micro explants cultured at POC have the best RNA quality and tissue structure whereas, macro explants were less viable after long-term culture. The findings also showed that placental explants undergoing syncytial degeneration produced poor quality RNA and should be avoided. The findings further showed that the RNA quality generically declined in tandem with STB degeneration, suggesting also that the STB viability could be a proxy measure of explant quality for high throughput experiments. In view of these, short term culture, macro explants and explants cultured in AOC were avoided in favour of micro explants in POC cultured at the gas interface to 6 days.

49 3.2.5 Modelling Defective Histamine-DAO-Axis (publications #9 - #11) Ex-vivo defective histamine – DAO – axis (dHDA) also referred to Elevated

Histamine Model (EHM) was modelled in human placental explants to identify genes that are affected by elevated histamine in the placenta and examined their functions in PE. Diamine oxidase (DAO; EC 1.4.3.22) also known as copper-containing-amine oxidase; histaminase; amiloride binding protein-1, was first discovered as a histamine-inactivating enzyme (Best 1929). This enzyme catalyzes the oxidative deamination of histamine to an aldehyde, ammonia, and hydrogen peroxide (Zeller

1938, Schayer 1959). DAO is mostly expressed in tissues (e.g., intestines, kidney, thymus, kidney cortex, seminal vesicles, placenta, and in the plasma of pregnant women), that transport large quantities of histamine (Marcou et al. 1938, Gunther &

Glick 1967, Tryding & Willert 1968, Beaven & Shaff 1975). And in publication #7, it was demonstrated that DAO is localized densely at the foeto-maternal interface to the cross borders between maternal (decidual) and foetal (trophoblastic) tissues, thereby confirming previous postulates about the topological orientation of the enzyme in human placenta (Buffoni 1966, Gunther & Glick 1967, Tornqvist 1969).

In normal human pregnancy, the parallel increase in DAO activity in the placenta and in maternal blood along with gestational age creates an effective histamine-DAO- axis (HDA), with DAO activity present throughout the course of pregnancy. In the pregnancies with normal HDA, maternal plasma DAO activity levels spontaneously rise exponentially to a 1000 fold from gestational week 8 and peaks at gestational weeks 20 and 24 (Ahlmark 1944, Southren et al. 1966b, Gunther & Glick 1967,

Weingold & Southren 1968, Tufvesson 1978, Beaven et al. 1975, Dubois et al.

1977). In contrast, the spontaneous exponential rise of DAO activity is abrogated from gestational week 8 onwards thus, leading to a defective histamine-DAO-axis in

50 pregnancies such as PE that present with high blood histamine (Southren et al.

1966b, Achari et al. 1971, Beaven et al. 1975, Legge & Duff 1981).

Thus, modelling the EHM was crucial to deciphering the functions of increasing histamine in PE. This was achieved by inhibiting DAO enzyme activity in regenerated syncytiotrophoblast placental explants with aminoguanidine to mimic in vivo condition. The explants were incubated at the liquid-gas interface in 8% oxygen,

37°C in basic media for 5 days (the culture media were changed at days 2 and 4) to allow for syncytiotrophoblast degeneration and regeneration. This was followed by a final media change on day 5, containing 100nM histamine in the presence of aminoguanidine (10-4 M final concentration) for 24 hours histamine treatment, or basic media without aminoguanidine for control samples. Endogenous placental

DAO metabolises histamine in 1g of placental explant at a rate of 45.54 µg/hr

(Kapeller-Adler 1952b). Therefore, to be able to observe the effect of the exogenous histamine, DAO activity was blocked with aminoguanidine in the treatment samples.

Aminoguanidine is a specific inhibitor of DAO enzyme activity (Tamura et al. 1989).

This culture method therefore enabled us to create an ex vivo elevated histamine model (EHM) with some parallel features of an in vivo defective histamine-DAO-axis system.

51 3.3 Sub-section 2: Justification for evolution of methodological techniques and contributions knowledge: - Transition from Single Gene to Multi-gene Analysis

Detailed analyses of biological processes that underpin health are identified by most

UK research funding bodies (e.g., MRC) as key step in the search for new ways of treating disease. Thus post-genome research focus is shifted to the accurate annotation of genomic sequences and the interplay between genes and proteins, and high throughput analysis of genetic variability between patients and susceptibility to disease or failed therapy. In tandem, methodological approach for the study of histamine role in placental gene expression evolved from a pre-genome approach focusing on single index genes using Polymerase Chain Reaction (PCR) technique through real time multiplex PCR to post genome high throughput techniques such as microarrays.

The changes to the technical approaches were necessary because of the multiple foci of the research agenda. First, a major focus of the study was to investigate the interplay between genes and proteins. Databases containing large numbers of genetic sequences and polymorphisms were available to facilitate the investigations however the method of investigation significantly influenced the rate of progress and dissemination. Similarly, to be able to significantly contribute towards faculty research output required maximisation of limited research time to decipher parallel gene expression profiles of normal and disease tissues in different clinical states.

Thus, there was an inherent need to maximise potentials and resources through technological advancement.

52 Cost effective analysis of the project in year Jan 2005 showed that it was extremely laborious, costly and time consuming to use low throughput techniques to investigate effect of histamine on placental gene expression. For example, in the year 2000, using conventional PCR technique, working 11 hours per day, five days a week for 6 months, enabled the study of the expression characteristic of 4 genes only. In contrast, the expression characteristics of 6 genes were studied over 3 weeks (7 hours per day; 3 days per week) in December 2004, with real-time multiplex qPCR technology at a fraction of the consumable cost for the year 2000 work. Moreover, the evolution of new high throughput technologies to quantify the expression of specific genes eliminated the need for some time-consuming and hazardous steps such as radio-labelling, scintillation counting, and post PCR analysis using carcinogenic chemicals such as ethidium bromide in gel electrophoresis. Thus on the grounds of cost effectiveness, expediency and safety, it was justified that application of high throughput technique to study histamine effects on placental gene expression was the best value for money option to significantly enhance the rate of data turnover and discoveries.

3.4 Sub-section 3: Microarrays and Bioinformatics: Considerations, Rationale and Contributions to Knowledge of Placental Gene Expression

3.4.1 Microarray Hybridisation (Publications #4, #9 and #11) In publications #4, #9 and #11, placental RNA hybridised onto Affymetrix microarrays were used to investigate the effects of oxygen and histamine in human placentae. In brief, RNA was extracted from micro explants with TRIZOL (Invitrogen, UK) and quality assessed. Total RNA was processed into labelled cDNA with NuGEN™

Ovation™ RNA Amplification System V2 and FL-Ovation™ cDNA Biotin Module V2

53 (Nugen). The resultant fragmented and labelled cDNA was added to the hybridisation cocktail in accordance with the NuGEN™ guidelines for microarray hybridisation onto Affymetrix GeneChip® Human Genome U133 Plus 2.0 arrays: one sample per array, (Figure 3-1) in Affymetrix GeneChip® Hybridisation Oven 640 for

18 hours at 45°C. Features that retained bound labelled cRNA after washing were visualized using the GeneChip® Scanner 3000 (Affymetrix).

Figure 3-1: Microarray Experiment Culture flowchart Total RNA from term placental explants were hybridised to Affymetrix Human Genome U133 Plus 2.0 Array. Explants were incubated for 6 days in POC or AOC. Half of the explants RNA under each of the ambient oxygen were treated with histamine and aminoguanidine in the last 24 hours of organ culture. Non-histamine & aminoguanidine treated samples served as the respective control samples.

3.4.2 Rationale for using Loop microarray design (Publications #4, #9 and #11) A critical component of microarray studies is to design the experiment to make the analysis of the data and the interpretation of the results as uncomplicated but as powerful as possible, given the purpose of the experiment and the constraints of the experimental material (Yang et al. 2002, Yang & Speed 2002). Thus, two main microarray experiment designs: single factor or one-way experiment, and multifactorial experimental designs (Yang & Speed 2002, Peng & Stromberg 2003) were

54 considered for the study to investigate the effects of histamine and oxygen in placental explants. Single factor designs typically include studies that investigate the differences in expression between two or more time points following treatment (time- course experiments), or different regions or subtypes of tissues such as normal and pre-eclamptic placentae (Kerr & Churchill 2001, Simon et al. 2002, Yang & Speed

2002, Blalock 2003). The microarray experimental designs that have two or more factors to be considered jointly, where each factor in turn may have two or more levels are normally considered multi-factorial (Yang & Speed 2002, Wit & McClure

2004).

Within the multi-factorial array designs, there are at least two basic types: a design with common reference or loop baselines (controls) (Kerr et al., 2000; Kerr &

Churchill, 2001b; Yang & Speed, 2002; Wit & McClure, 2004; Kerr & Churchill,

2007). A loop design (Figure 3-2a) describes the method where samples are directly compared by direct pairing of sample array slides, usually based on scientific aim or biological interest of the study, whereas a common reference design describes the method where all samples are indirectly compared by referring to a single reference sample (Figure 3-2b) usually a pre-culture time-zero sample (Kerr & Churchill,

2001a; Yang & Speed, 2002; Wit & McClure, 2004). In effect, loop design uses direct comparison between slides without a common reference so that the choice of a baseline control depends on the experimental question.

Figure 3-2: Microarray Experiment Design

Previous comparative work reporting on the appropriateness of loop design (Yang &

Speed 2002) showed that loop design gave the smallest confounding variance for estimating differential expression between the main effects, whereas common reference usually derived from time zero samples (e.g., Figure 3-2b), by far gave the worst variance for estimating the interaction in both 2 X 2 factorial and single factor time-course experiments (Yang & Speed 2002). Thus, in our loop design (Figure

3-2a), samples were compared with each other in a loop order according the biological or experimental questions. Our direct loop order was driven by 3 main scientific questions aimed at determining whether: (1) gene expression between explants cultured for 6 days in POC or AOC differ, (2) the effect of histamine on placental explants gene expression differ between those cultured in POC and AOC; and (3) gene expression in cultured explants with diminished DAO activity but treated with histamine differ from cultured explants without diminished DAO and histamine treatment. Therefore the loop design allowed for direct comparison of time equivalent controls with treated sample. For example, in order to determine relative

56 changes in gene expression between explants cultured in POC and AOC, slide A was compared with B where slide A was made the control (Figure 3-2A). Therefore, based on the biological interests posed by the experimental questions a loop design was deemed more appropriate to enhance the biological relevance of the results.

3.4.3 Dealing with the impacts of variances (Publications #4, #9 and #11) While it is well demonstrated that one of the best designs to limit confounding variance in multi-factorial microarray experiment is loop design instead of common reference design (Kerr et al., 2000; Kerr & Churchill, 2001b; Yang & Speed, 2002;

Wit & McClure, 2004; Kerr & Churchill, 2007), there are constraints associated with this type of experimental designs. Often, the constraints are linked to strategies to deal with inherent variances that may affect data interpretation. In this instance, the documented sources of variance have broadly been classified either as biological or technical variance (Blalock 2003). At the beginning of the studies, it was clear that both biological and technical variances in microarray studies could be effectively contained with judicious statistically based experimental design to account for four basic factors: control, randomisation, replicates and balance (Peng & Stromberg

2003, Schena 2003, Wit & McClure 2004).

Replicates in microarray studies are deemed highly important in creating a degree of variation in measurements to provide independence of data, and to offer the benefits of averaging log expressions to infer external validity (Lee et al. 2000, Yang & Speed

2002, Peng & Stromberg 2003). The two types of replicates (biological and technical) described in microarray experiments appear to have different degrees of impact on variance in gene expression. The technical replicate refers to replication in which the

57 target mRNA is from the same pool or extraction, whereas biological replicate refers to hybridizations that involve mRNA from different extractions or different samples of cells from a particular cell line or tissue; or from target mRNA, which comes from different individuals or different versions of a cell line (Lee et al. 2000, Yang & Speed

2002).

Of these, it is suggested that technical replicates generally involve a smaller degree of variation in measurements than the biological replicates, and therefore do not provide a greater deal of impact on external validity (Lee et al. 2000, Yang & Speed

2002, Peng & Stromberg 2003). In contrast, the literature suggests that external validity of microarray findings can be improved with meticulous prediction of biological replicates to ascertain appropriate sample size to detect statistical differences in gene expression along with appropriate use of control samples to establish baseline expression (Yang & Speed 2002, Wright & Simon 2003, Dobbin &

Simon 2005). Analysis of variance distribution incorporating random variance model to estimate the within-class variance for each gene in the human genome to predict sample size (Wright & Simon 2003, Dobbin & Simon 2005) was therefore performed in Chipster v1.4.7. Consequently, biological replicates of 6 per class was predicted to identify differentially expressed genes at Type 1 error = 0.05, (Power = 70%, hypothesized mean difference in log2 expression between classes = 1).

58 3.4.4 Validation of the Microarray Design: Biological Relevance of Pre-culture explant as baseline control for histamine treated cultured explants (Publication #3 & #4) To validate the loop design, the effects of tissue culture and oxygen on placental gene expression were examined to ascertain the biological relevance of studying histamine effects in POC or AOC, and also for using pre-culture explants as control for the cultured explants. It was argued that pre-culture (time zero) placental explant expression was needed as a common reference baseline control to put the cultured explant microarray expression into perspective as opposed to using cultured non- treated explants as baseline for the cultured treated samples. Therefore, the hypotheses that (1) entropy of gene expression of pre-culture time zero explant and explants that have been cultured and undergone STB regeneration were mutually meaningful and biologically relevant; (2) oxygen concentration and tissue culture per se do not affect the explant gene expression and the interpretations thereof were tested.

3.4.4.1 Biological Relevance between Pre-culture and cultured Explants for RGE (#4, 9, 11) To answer the question of using time zero explants as common reference baseline for the cultured explants, biological relevance network analysis was performed with placental micro explants cultured for 6 days in POC (4 biological replicate arrays), in

AOC (6 biological replicate arrays) and pre-culture time zero explant (5 biological replicates) microarrays. All placentae were delivered by elective caesarean sections.

The biological relevant networks analysis was performed by computing the entropy of gene expression patterns between the microarray samples to provide mutual information (information content) for RNA expression patterns from each hybridised explant sample (Butte & Kohane, 2000). The correlation coefficient between explants was measured by comparing the expression pattern of genes on each slide to that of

59 every other gene. Minimum Pearson correlation (R2) threshold was set as

0.7700514, and maximum Pearson correlation (R2) threshold at 1.0. To observe all true relevance networks, highest entropy filter was also set to 100%.

The analysis showed 55 biologically relevant clusters or links (Figure 3-3), with two mutually exclusive relevance subnets. Subnet 1 contained 10 experiments entirely from explants cultured for 6 days, while subnet 2 contained 5 experiments exclusively from pre-culture (time zero) explants. The percentages of biologically relevant networks in each subnet were respectively 67% and 33% for subnet 1 and

2. The explants were represented as nodes in the network and lines were drawn between the nodes (Figure 3-3) to show the link was biologically relevant when the respective correlation coefficient was R2 >0.77. All gene expression entropies for the

15 microarrays were linked through comprehensive pair-wise mutual information analysis, but no biologically relevant network was identified between pre-culture time zero explants and 6 days cultured explants. The findings suggested that the links between gene expression in pre-culture time-zero and explants cultured to 6 days were not biologically relevant to support direct comparison. Thus, indicating that the use of pre-culture time-zero explants as a baseline control for 6 days culture time- point placental microarray experiment could complicate interpretation of results.

Figure 3-3: Biological Relevance Network Analysis of Time zero, POC and AOC treated Explants Fifteen placental explant microarrays from 11 elective caesarean normal placentae (pre-culture explants = red nodes, n = 5; explants cultured in POC for 6 days = green nodes, n = 4; explants cultured in AOC for 6 days = blue nodes, n = 6) were analysed to identify biologically relevant networks (BRN) between each pair of samples. A total of 55 BRNs were identified. (a) = true links between cultured explants (green nodes = POC explants; blue nodes = AOC explants); (b) true links between pre-culture explants (orange nodes). No BRN were found between pre-culture and cultured explants at minimum Pearson Correlation R2 = 0.77. Networks (lines) coloured in red represent elements that are positively correlated while networks coloured in blue represent elements that are negatively correlated.

3.4.4.2 Impact of Tissue Culture and Oxygen on Explant Gene Expression (#4) To further examine the implication of the lack of biologically relevant links between pre-culture and 6 days cultured explants on the interpretation of results, the differences in placental transcriptomic changes in response to tissue culture and oxygen, and the relevance of genetic alterations and pathways associated with POC and AOC culture were investigated. Placental explants were collected from normal term (38-39 weeks of gestation) placentae with no previous uterine contractile activity. Placental gene expression profile was evaluated with analysis of GeneChip®

Human Genome U133 Plus 2.0 arrays (Affymetrix).

61 Absolute gene expression (AGE) analysis was performed using One Class

RankProd statistics to identify ‘consistently high expressed’ (CHE) and ‘consistently low expressed’ (CLE) significant genes respectively in Time Zero (pre-culture) explants (designated as T0), and explants cultured in POC and AOC for 6 days. FDR

Confidence (1-alpha) was set at 99.9% (FDR <0.001). The analysis identified a total of 635, 1207 and 1760 significant genes that were consistently expressed in pre- culture (T0); POC and AOC cultured explants respectively (Figure 3-4). Of these,

224 genes were exclusively expressed in pre-culture samples. In contrast, 69 genes

(1 CHE and 68 CLE) were exclusively expressed in POC and 574 genes (292 CHE, and 282 CLE) in AOC cultured explants only.

The analysis also identified a further 48 genes that were expressed in both pre- culture and AOC treated explants but not in POC treated explants. In addition, it was observed that the CHE genes appeared insensitive to POC, but rather to tissue culture and AOC. In that, all CHE genes (except SIGLEC6) expressed in POC explants were also expressed consistently at high levels in AOC explants.

Furthermore, there was a core set of 770 significant genes that were expressed in both POC and AOC treated samples only, and appeared to be genes suggestively responding to tissue culture per se. No genes were exclusively expressed consistently between pre-culture and POC treated samples only. However, using class prediction modelling incorporating Leave-one-out cross-validation and ROC curves analysis, the study identified the expression of 134 genes that were significantly (p<0.01, AUC = 0.82(CCP), 0.81(DLDA), 0.81(BCCP)) associated with prolonged explant culture at AOC.

Figure 3-4: Expressed genes in POC and AOC explants Figure A shows overlap of consistent high level expressed (CHE) genes between pre-culture, and explants culture at AOC and POC. Figure B shows overlap between pre-culture CHE genes and cultured explants consistent low expressed (CLE) genes. Figure C shows a volcano plot of 157 significant genes (blue dots). Random variance model parameters at a= 1.31652, b= 17.03489, Kolmogorov-Smirnov statistic= 0.01 and a nominal significance level (dotted line) of each univariate test at p< 0.05 (210 exact permutations). Figure D shows the ROC curve from the Bayesian Compound Covariate Predictor for AOC associated genes. S1 – S8 = Gene Sub-sets 1 – 8; T0 = Pre-culture CHE genes; AOC = Atmospheric Oxygen Concentration; POC = Physiologic Oxygen Concentration; +ve = CHE; -ve = CLE.

Subsequently, through analysis of the gene ontologies and pathways enriched by the pre-culture, POC and AOC significant gene sets it was shown that the preserved set of genes that appeared to respond to tissue culture irrespective of oxygen concentration were significantly involved in programmed cell death, cell death, death, stress response, protein metabolic process, electron transport activity, RNA

63 translation and oxidoreductase activity, regulation of cytosol, cytoplasmic and organelle parts of the placental cells. These therefore suggested that tissue culture per se induces the expression of genes that regulate a cluster of cathartic functions in the placenta.

The work was thus extended to examine the effects of culture oxygen concentration on Transcription Factor Target gene sets (TFTs) expression. It was observed that culture at AOC down-regulated specific TFTs including REL, ETV4, ATF3, STAT1,

JUND and STAT5B. Through text mining analysis, these TFTs were identified to be involved in cell growth, proliferation, invasion, regeneration, differentiation, transformation, tissue viability, protection from apoptosis, and glands development.

Conversely, AOC up-regulated TFTs such as HIF1A, PPARA, CEBPD, STAT3, and

CEBPE, that are associated with oxygen regulation, immune and inflammation responses; leptin mediated response; lipid metabolism; suppression and mis- regulation of cell growth and proliferation; cell morphogenesis; induction of apoptosis; and oxygen regulation. These findings collectively suggested that:

1. Gene expression patterns differ between pre-culture and cultured explants

2. Placental explants’ responses to tissue culture per se induce expression of

genes with preponderance towards apoptosis, inflammation and tissue

response to stress irrespective of culture oxygen concentration.

3. The magnitude of the tissue response to stress and programmed cell death in

placenta is sensitive to AOC.

4. Explants culture at AOC exacerbates tissue response to stress and

programmed cell death in placenta.

64 5. AOC up-regulates HIF1A transcription target gene set

6. Oxygen and tissue culture generally precipitate down-regulation of a sub-set

of placental genes

Overall, the findings suggested that the lack of biologically relevant links between pre-culture and 6 days cultured explants was underpinned by critical biological variance, and that the use of pre-culture time zero explants as baseline for 6 days cultured explants could complicate the interpretation rather than offer meaningful baseline. The findings also confirmed previous reports (Thompson et al. 2007,

Fajardy et al. 2009) that culture per se affects explant gene expression. It further showed that irrespective of oxygen concentration, tissue culture per se down- regulates more genes but the effect is exacerbated in explants cultured under AOC.

Therefore, it was acknowledged that in a time point (not time series) experiment variances introduced by explant culture including syncytial degeneration and regeneration and differential gene expression would complicate the interpretation of gene expression induced by histamine after long-term culture, and hence pre-culture explant was deemed unsuitable as baseline control for relative gene expression analysis (RGE).

3.4.5 Rationale for Bioinformatics methodologies (Publications #2, #4 and #11) Bioinformatics analyses were performed to transform gene expression data into biological information. R Bioconductor statistics were used within R Studio (RStudio

2015), BRB ArrayTools (Simon et al 2007), Gene Ontology Tree Machine (GOTM,

(Zhang et al. 2004)), "WEB-based GEne SeT AnaLysis Toolkit (WebGestalt, (Wang et al. 2013)), MEV4 (Saeed et al. 2006), Reactome (Fabregat et al. 2016), Chipster

(Tuimala et al. 2008), INMEX (Xia et al. 2013) and NetworkAnalyst (Xia et al. 2015)

65 to process microarrays and also to probe the Gene Ontology, Kyoto Encyclopedia of

Genes and Genomes (KEGG), BIOCARTA and Molecular Signature Database

(MSigDB) data bases to assess the biological relevance and significance of genes according to similarity in function, participation in signalling pathways, and protein domains.

3.4.5.1 Gene Findings: AGE and RGE (#2, #4, #11) The usual approach to finding significantly expressed genes from microarray experiments relies on relative gene expression (RGE), which is defined as the relative quantitation of the differences in the expression level of a gene between the case and control samples (Xia et al. 2013). Traditionally, this approach is suggested as highly suitable for candidate gene discovery or class prediction studies, but limited in its ability to find genes that are significantly expressed consistently at low or high levels in the same class or phenotype (Schena et al. 1995, Dudley et al.

2002, Simon et al. 2002). To find such regulated genes would require absolute gene expression (AGE) analysis, but existing methods for microarray AGE relied on common reference methods (Seita et al. 2012). In the common reference method, microarray data were normalized individually with the common reference and averaged. These averaged values were then mapped onto the probeset meta-profile in order to obtain the population’s percentile and expression value for each probeset

(Seita et al. 2012). While this methodology was helpful if the objective was to compare a probeset value against existing values, the value thus computed was relative and arbitrary; and subject to changes to the pooled value. It was therefore proposed that RankProd analysis method could be used to perform AGE without reliance on a common reference.

RankProd analysis is a non-parametric statistic originally modelled for RGE (Breitling et al. 2004). It combines the gene rank from the different arrays together instead of using actual expression data to select genes that are consistently ranked high or low

(Breitling et al. 2004, Hong & Breitling 2008). The product ranks from all samples are then calculated as the test statistic in100X permutations with a nominated False

Discovery Rate (FDR), for example, confidence at 1 – alpha = 95.0% or 99.9%. The

RankProd analysis generates three pools of data classified as Negative Significant,

Positive Significant and Non-significant. In the context of intra-class AGE, negative significant gene pool contains genes that are consistently expressed at low levels

(CLE) across the libraries from the same class, and are thus classified as down- regulated (Saeed et al. 2003). Positive significant gene pool on the other hand contains genes that are consistently expressed at high levels (CHE) across all libraries from the same class (up-regulated) (Saeed et al. 2003). Genes that have very low expression below the present call or background threshold cut-off point, and genes that have inconsistent expression across all libraries from same class are classified as non-significant (Saeed et al. 2003).

In contrast, positive significant genes (up-regulated) identified from RGE are genes that are more highly expressed in treatment group relative to the expression in control class or libraries; and negative significant genes (down-regulated) are the genes that are more highly expressed in the control class relative to treatment class.

Both RGE and AGE methods have been used at different stages in the work to answer different questions. For example, in paper #2, AGE methodology was for the first time used successfully to identify comprehensive gene-signature that included

67 genes with consistently low or high expression in both pre-eclamptic and normal placentae. The findings showed that the use of AGE analysis enables independent description of a comprehensive, globally and consistently expressed genes in the case and control samples. On the other hand, the findings showed that overt use of

RGE analysis to the disadvantage of AGE could limit gene sets and our understanding of the real time and complexities of changes that could occur in a case or disease state.

The findings also confirmed earlier reports (Simon et al. 2003, Breitling et al. 2004) that RGE not only identifies limited candidate genes but could also exclude large proportion of genes that may be of relevance in characterising the molecular pathology of a disease including those with biologically relevant low level expression and genes with similar levels of expression in both the case and control samples.

The findings further suggested that RGE could inherently identify genes whose expression patterns may be inconsistent but might have large differential expression between control and case samples.

3.4.5.2 Finding Biological Relevance of Genes (#2, #4 & #11) For interpreting the expression results in terms of higher order of biological processes or molecular pathways, Gene Set Analysis methodology (GSA) method

(Efron & Tibshirani 2007), Gene Set Enrichment Analysis ((GSEA), (Subramanian et al. 2005)), Over-representation analysis, Paroto Analysis for Vital Few Genes

and Leading Edge Metagene ((LEM), (Subramanian et al. 2007)) analysis were used to identify over-represented, enriched and leading edge genes from Gene

Ontology (GO) term, genetic pathways, transcription factor target gene sets.

68 3.4.5.3 GSA and GSEA (#2, #4 & #11) GSEA is a computational method that determines whether a priori defined set of genes shows statistically significant, concordant differences between two biological states (e.g. phenotypes) (Subramanian et al. 2005). GSA was mostly used in the studies that focused on deciphering the differential enrichment of pathways between experimental and control samples. The Efron-Tibshirani's GSA uses the maxmean statistics, a modification of the GSEA procedure (Subramanian et al. 2005) for assessing significance and direction of regulation of pre-defined gene-sets. The maxmean statistic identifies the direction of regulation of the gene set by a measure of the absolute value of the positive or negative part of gene scores di in the gene set (Efron & Tibshirani 2007). In this analysis, all gene scores di in the gene set were taken to set all of the negative ones to zero. Average of the positive scores were then taken to give a positive part average = avpos. Similarly, gene ‘scores di’ in the gene set were taken to set all of the positive ones to zero and then average for the negative side were taken to give the negative part average avneg (Efron & Tibshirani

2007, Simon et al. 2007). Thus, the overall score for the gene set was avpos (up- regulated) if |avpos| > |avneg|, or otherwise it is avneg (down-regulated) (Simon et al. 2007). Consequently, transcription factor target gene sets and biochemical pathways analyses were performed.

3.4.5.4 Leading Edge Metagene (#11) LEM is build-up of the GSEA to find the leading-edge subset in a gene set. Leading- edge subset in a gene set are those genes that appear in a ranked list at or before the point at which the running sum reaches its maximum deviation from zero

(Subramanian et al. 2005). LEM is useful for identifying the core genes that account

69 for the gene set’s enrichment signal. Therefore in the investigation of the effects of histamine on placental gene expression, downstream analysis of the gene set enrichment results with GSEA leading edge metagene (LEM) method was performed to identify groups of genes that were co-regulated by histamine, and are likely to form the core sub-signatures of genes that accounts for the gene set’s enrichment signal to establish the distinct biological processes or pathways in PE placentae.

3.4.5.5 Pareto Analysis for Vital Few Genes (#11) The Pareto Principle was applied to identify the "vital few" genes responsible for producing most of histamine effects. The principle as originally applied to quality improvement suggests that a great majority of effects (80%) are produced by a few key causes (20%) (Juran 1975). The importance of identifying the vital few histamine regulated genes lies in the plausibility that nothing of significance can happen in response to histamine unless these vital few genes are regulated. The analysis involved rank ordering of the frequency of involvement of a particular gene in multiple biological functions or ontologies, and applied the Lorenz (Lorenz 1905) cumulative curve to depict the distribution of the vital few genes.

3.5 Conclusion There were many challenges for studying histamine effects on the placenta. The challenges ranged from using an appropriate ex vivo model that limited iatrogenic inflammation to tissues that are primed for or are already pro-inflammatory to maintaining maximal concentration of histamine introduced to experimental samples.

These experimentations made substantial contributions to these areas of placental

70 research by showing that inflammatory effects can be controlled to investigate the effects of histamine on placenta that have undergone Syncytiotrophoblast degeneration and regeneration. These studies also showed that it is possible to obtain high quality RNA akin to pre-culture quality from micro explants cultured for 6 days at the liquid-gas interface. These studies also confirmed that placental explants are viable after 6 days culture but the explant size affects the viability and integrity after culture. These studies further showed that morphological assessment of the degree of STB degeneration is parallel to explant RIN and could be used as proxy measure for explant viability. These investigations further contributed to the knowledge of placental tissue culture by providing genomic evidence to confirm that placental explant culture at atmospheric oxygen was pathological and that culture per se triggered the expression of cathartic genes, and the effect is worsened by culture in atmospheric oxygen. In contrast, explant culture in an optimum hypoxic condition (8% oxygen) was physiological and optimum for term placental explant culture. The discussion of the rationale underpinning the bioinformatics approach that incorporated absolute gene expression modelling to uncover PE pathways and the complexity of histamine pathways in PE placentae demonstrated additions to knowledge about novel methods to architype intra-phenotype significant gene expressions.

71 Chapter 4 Theme 3: Understanding causes of elevated Histamine in Pregnancy and Pre-eclampsia

In this chapter: The commentary is based on publication items, #5, #6 and #7:  (#5) Brew O, Lakasing L, Sullivan M. (2007) Differential Activity of Histidine Decarboxylase in Normal and Pre-eclamptic Placentae. Placenta. May-Jun; 28(5-6):585-7.

 (#6) Brew O, Sullivan, M. H., & Roller S 2005, "Regulatory loops between cytokines and histamine in the human placenta", Placenta, vol. 26, no. 8-9, p. A.52

 (#7) Brew, OB & Sullivan MH (2001) Localisation of mRNAs for diamine oxidase and histamine receptors H1 and H2, at the foeto-maternal interface of human pregnancy, Inflammation Research, 50(9): 449-452

The focus of the commentary is to discuss the investigations that broadened our understanding of the causes of elevated histamine in pregnancy, with emphasis on histamine synthesis, production and metabolism in PE placenta.

Highlights:

 Histamine synthesis based on de novo HDC activity is higher in PE placentae than in normal placentae

 The cytokines IL-10, IL-1β and INF- time dependently regulated histamine production in human placenta, inducing both rapid (30 minute) and prolonged (24h) increases in histamine.  Th1 (INF-) and Th2 (IL-10) like cytokines, pro-inflammatory (IL-1β) cytokines

and endotoxins induce biphasic histamine increase in placenta. The responses are suggestive of both histamine release from storage and de novo histamine synthesis.  Histamine receptors are expressed in human placenta.

 The co-expression of histamine receptors and DAO is consistent with a role for histamine at the foeto-maternal interface of human pregnancy  There was a regulatory feedback loop between histamine, cytokines and DAO expression in human placenta.

 The expression of DAO in the decidua is widespread, indicating that the enzyme could also function to prevent prolonged exposure of decidual cells to bioactive histamine.

72 4.1 Introduction:

PE is a complex disorder of pregnancy that involves a systemic inflammatory response, endothelial activation within the maternal vascular system, increased maternal blood histamine and decreased histamine metabolism (Kapeller-Adler

1941, Redman et al. 1999, Chambers et al. 2001, Sargent et al. 2003). A number of the features of PE (hypertension, proteinuria, oedema, nausea, and headaches) as outlined in publication item #1 also mimic those of hyperhistaminemia (hypertension, proteinuria, oedema, nausea and vomiting, headaches, tachycardia, palpitation, convulsions, metallic taste). Normally, maternal plasma DAO levels increase up to

1000-fold during pregnancy (Southren et al. 1966a, Beaven et al. 1975), but this increase in DAO is attenuated in PE, with a resultant high blood histamine (Kapeller-

Adler 1944, Achari et al. 1971).

The placenta is the primary source of histamine and DAO in pregnancy. However, the human placenta functions both to produce nutrition and remove waste products for the developing foetus (Lyall 2002), thus suggesting that histamine as found in the placenta with increased levels in PE placentae could be a mere by-product of foetal development and hence may not have functional roles in the placenta. Therefore it was hypothesised that (1) for histamine to have functional roles in the placenta it must express receptors at the foeto-maternal interface; (2) that placental HDC activity did differ between normal and PE tissues; and (3) that pro-inflammatory cytokines and endotoxins would increase histamine synthesis in the placenta.

Using DNA and RNA extraction, gene cloning, gene amplification and in situ hybridization technologies, the expression of mRNAs for histamine receptors H1 and

73 H2, and for diamine oxidase at the foeto-maternal interface were localised. Similarly, using enzyme assays including fluorometric and chemiluminescent methodologies and RNA amplification with reverse transcription polymerase chain reactions (RT-

PCR) HDC activity and the effect of histamine on cytokines expression in the placenta were investigated. The commentary will critically reflect on the relative contribution made by these findings to the knowledge and understanding of molecular functions of histamine in human placenta.

4.2 HDC activity in PE placentae (Publication #5)

In theme 1, it was established that maternal histamine levels in PE increases above levels found in normal pregnancies, and the difference is significant (Figure 2-6). It was also established that the placenta is the source of the elevated histamine in human pregnancy and that nascent or de novo histamine rather than pre-formed histamine is the most likely source of the elevated histamine in PE. In order to increase our understanding of the source of the elevated histamine in PE, an enzyme assay study using fluorometric, ELISA, and chemiluminescent methodologies was conducted to test the hypothesis that placental HDC activity did not differ between normal and PE tissues and histamine.

The findings showed that HDC activity was present in all placental tissues tested, and normal placentae produced histamine at the rate of 4.77 ± 0.87 ng/g/hr, compared with 5.49 ± 0.52 ng/g/hr from the pre-eclamptic tissue (p<0.05, ANOVA).

The work also showed that histamine production varied substantially between tissues, so the net rate of production for each placenta (Histamine level after 1 hour

74 of incubation – histamine level in corresponding control tissue) was compared using log ratios. In normal tissues this was -0.37 ± 0.26 ng/g/hr, compared with 1.86 ± 0.5 ng/g/hr from pre-eclamptic tissue (means ± s.e.m.), a significant difference (p =

0.0003 ANOVA). Thus confirming that histamine synthesis based on de novo HDC activity is higher in PE placentae than in normal placentae, and could contribute to the increased levels of histamine in maternal blood in pre-eclamptic pregnancies.

This finding did not identify whether the change in pre-eclamptic placental histamine is a cause of the disorder, or an effect of it, but it made it clear that the placenta can contribute to the pathology. Although this study showed that the rate of histamine synthesis is increased in PE placentae, it could not confirm previous report that histamine content may be increased in pre-eclamptic placentae (Szukiewicz et al.,

1999a). This observation was congruent because the study was intended to investigate synthesis rather than storage which may explain the difference. The findings however confirmed that the placenta contains histamine (Kapeller-Adler,

1952; Purcell, 1992), which would be produced by HDC. It was therefore suggested that in normal pregnancy histamine is stored in the placenta but the synthetic capacity may be relatively low compared to pre-eclamptic placentae.

75 4.3 Effects of Cytokines and Endotoxins on Histamine Production in the Placenta (Publication #6)

In attempts to investigate the causes of elevated histamine in PE, the effects of index cytokines for inflammation and endotoxins on histamine production in the placenta were examined. In this work, a total of six placental samples were collected from 6 different normal pregnancies. Three placental explants per sample were used to investigate the effect of key cytokines and endotoxins on histamine production in the placenta. Index cytokines Interleukin 1 beta (IL-1), Interferon gamma (IFN-),

Interleukin 10 (IL-10) respectively for pro-labour, Th-1, Th-2 cytokines families and endotoxin lipopolysaccharide (LPS) were tested.

Tissues in triplicates were stimulated with IL-1 (1 ng/ml), LPS (10 ng/ml), IL-10 (10 ng/ml), IFN- (100 U/ml) in the presence of aminoguanidine (10-4 M final conc.) and cultured to different time courses (30min, 1hr, 2hr, 4hr, 8hr & 24hr). Histamine was then measured with histamine ELISA (IBL, Hamburg) and corrected for tissue weight. The difference in histamine production in response to the stimulant over the specified time course were statistically analysed with ANOVA and Bonferroni/Dunn post hoc to determine significance difference. The analysis showed that histamine was produced with a mean baseline (time zero) value of 31.1± 4.4 nmol/L (mean ±

SEM, n = 6), and IL-10, IL-1beta and INF- time dependently regulated histamine production in human placenta (Figure 4-1).

Figure 4-1: Cytokines and Endotoxin effect on Placental Histamine Production Figures A – D show the effects of the index cytokines Interleukin 1 beta (IL-1), Interferon gamma (IFN- = INF-gama), Interleukin 10 (IL-10) and Lipopolysaccharides (LPS) on histamine production in normal placenta respectively at times 30 mins, 1, hour, 2 hours and 4 hours. * = significant difference between control and stimulant induced histamine levels.

Figure 4-1A – D showed that IL-1, IFN-, IL-10 and LPS up-regulated short term histamine production in human placenta. All stimulants used significantly (overall p<0.0001 Bonferroni/Dun test) increased histamine levels after 30 minutes of culture

(Figure 4-1A). DAO activity was blocked during culture. This implies histamine levels would naturally rise during culture. Therefore stimulant induced histamine production was compared with time-matched control (non-treated) samples. While the effects of the endotoxin (LPS), pro-labour (IL-1,) and Th-1 cytokines (IFN-) on histamine production were transient with maximal response at 2 hours (Figure 4-1C), the effect

77 of the Th-2 cytokine (IL-10) was sustained at the end of the 4 hours of culture (Figure

4-1D).

The transient effect observed at the end of 4 hours was reversed after 24 hours

culture (Figure 4-2). All cytokines and the endotoxin tested significantly increased

histamine levels after 24 hours culture (Figure 4-2). Thus, indicating that the index

cytokines and endotoxin could both release histamine from storage (effect at 30

minutes) and induce histamine synthesis (effect after 24 hours).

p = 0.0027 p =0.0088

Figure 4-2: Effects of Cytokines and Endotoxin on Placental histamine production after 24 hours of culture Figure shows the effects of the index cytokines Interleukin 1 beta (IL-1), Interferon gamma (IFN- = INF-gama), Interleukin 10 (IL-10) and Lipopolysaccharides (LPS) on histamine production in normal placenta after 24 hours. * = significant difference between control and stimulant induced histamine levels.

These results showed that IL-1, IL10 and LPS overall increased histamine levels at

the 0.1 µmolar range, and that in the short term the pro-labour cytokines, Th-2 like

cytokines and endotoxins could stimulate histamine production in the placenta at

concentrations similar to levels expected to mediate physiological functions such as

78 such tissue remodelling, and cell proliferation while the INF  (TH-1 cytokine index) produced histamine in the range normally associated with pathophysiology.

4.4 Histamine receptors and DAO mRNA are expressed in human placenta (Publication #7)

To determine whether the histamine thus produced in the placenta has functional roles in the placenta, DNA and RNA extraction, gene cloning, gene amplification and in situ hybridization technologies were used to examine and localise the expression of mRNAs for histamine receptors H1 and H2, and for diamine oxidase at the foeto- maternal interface. The results showed that co-expression of histamine receptors and DAO is consistent with a role for histamine at the foeto-maternal interface of human pregnancy.

These findings are important, not only for the understanding of possible role of histamine in human pregnancy but also showed for the first time that H1R, H2R and

DAO mRNAs are expressed in juxtaposed positions in the placental and decidual components of the foeto-maternal interface of human pregnancy. The expression of mRNAs for histamine receptors H1 and H2 in human amnion, chorion, decidua and villous cytotrophoblast and stromal cells was novel. Previous studies at the time of publication had reported on finding the genes for histamine receptors 1 and 2 in human placenta tissues (Murakami et al. 1999) but not on the expression of the genes for the receptors in the specific tissues. Subsequent studies built upon this work to examine the expression of receptors in the syncytiotrophoblast cell layers

(Matsuyama et al. 2004, Matsuyama et al. 2006).

79 The presence of mRNA for DAO in human placenta was documented previously

(Zhang et al., 1995; Zhang & McIntire, 1996), but the localisation of mRNA expression had not been previously reported at the time of going to press. Therefore this work was significant in that it extended the understanding and knowledge of the expression site of the enzyme in the syncytium. The work further provided insight into the plausible role of DAO as a metabolic barrier to prevent excessive entry of active histamine into the maternal or foetal circulations. Similarly, the widespread expression of the enzyme in the decidua showed that it could also function to prevent prolonged exposure of decidual cells to bioactive histamine.

4.5 Conclusion

As the expression of histamine receptors is generally required for any functional effects of the amine in tissues, the localisation of histamine receptors in human amnion and chorion as well as in decidua and villous trophoblast cells was a crucial discovery. The findings clearly showed that these tissues could be targets for the histamine thus produced de novo in the placenta, and therefore enhances our knowledge and understanding of the molecular functions of histamine in human placenta. These studies further showed that products of inflammation, tissue remodelling, cell proliferation as normal placentation processes could trigger increased production of histamine. The findings therefore demonstrate the contribution made to the knowledge and understanding of the mechanisms that may underlie the regulation of histamine production in human placenta; a breakdown of which could contribute to elevated histamine associated with pre-eclampsia.

80 Chapter 5 Theme 4: Effects of histamine on placental genes expression: Implications for pre-eclampsia

In this chapter:

The commentary is based on the publication items #2, #6, #8 - #11

 (#2) Brew, O., Sullivan, M.H. & Woodman, A. (2016) Comparison of Normal and Pre -Eclamptic Placental Gene Expression: A Systematic Review with Meta-

Analysis. PloS One, 11(8), p.e0161504.  (#6) Brew O, Sullivan, M. H., & Roller S (2005), "Regulatory loops between cytokines and histamine in the human placenta", Placenta, vol. 26, no. 8-9, p. A.52  (#8 ) Brew, O. & Sullivan, M.H.F. (2007) Histamine regulates placental diamine oxidase mrna expression - Evidence for a feedback loop decreasing histamine production in pregnancy? Placenta, 28(8 - 9), p.A71.  (#9 ) Brew OB, Sullivan MHF (2015) Placental Gene Expression in Response to Histamine and Oxygen, Gene Expression Omnibus. Series accession number

GSE74446 (http://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE74446)

 (#10 ) Brew O, Sullivan M. (2017) Elevated Histamine Model: A Protocol for an ex vivo model for in vitro study of histamine effect on placenta. Protocols.io dx.doi.org/10.17504/protocols.io. jigckbw

 (#11) Brew O, Sullivan MHF. Comparison of elevated histamine model and pre- eclamptic placental gene expression,

Highlights:

 Histamine and DAO: two major biomarkers implicated in pre-eclampsia form a bi- directio nal regulatory negative feed-forward relationship in the placenta.  Increasing histamine down-regulates DAO mRNA expression and this could lead to

diminished DAO activity in the placenta.

 Histamine up-regulated IL-1b, IL-10 and INF- production term placentae and this was reflected in increased long-term changes in mRNA levels.  The relative up-regulation of Th-1 cytokine INF- in response to prolonged exposure of

high dose histamine is greater than that of IL-10 (Th-2 type cytokine).  Pre-eclampti c (PE) placentae express unique set of genes that are not significantly expressed in Normal (NP), but retain its ability to express the genes associated with in

NP.  NP and PE placentae both share common biological pathways but additional pathways that are not significantly affected in NP are uniquely enriched in PE placentae.

 Multiple Over-Represented genetic pathways underpin the aetiology of PE.

 The natural process by which pre-eclampsia develops is affected by the amount of histamine in the maternal blood and placenta.  The natural process by which pre-eclampsia develops is indirectly affected by histamine

via covariate functional groups regulated by specific PE placental genes.

81 5.1 Introduction Several studies have demonstrated that histamine signalling through HRH1

2+ generally activates cGMP, Ca fluxes, CaMKII, NFκB, Phospholipase A2, and

Phospholipase D (Hill 1990, Jutel et al. 2001, Bakker et al. 2004, Matsubara et al.

2006). In contrast, HRH2 activates cAMP, PKC- PKC-1 adenylate cyclase, c-Fos, c-Jun, and Ca2+ fluxes in the placenta (Beaven 1982, McMillan et al. 1985, Hill 1990,

Ghosh et al. 2001, Jutel et al. 2001, Davio et al. 2002). These histamine signalling pathways have been linked to regulation of placental cytokines, trophoblast proliferation, differentiation, inflammation, placental growth and cervical remodelling

(Falus & Meretey 1992, Hamano et al. 1998, Jutel et al. 2001, Ma et al. 2002, Liu et al. 2004, Cricco et al. 2006, Matsubara et al. 2006, Matsuyama et al. 2006, Fukuda et al. 2007). Nonetheless, knowledge on placental gene expression in response to histamine was limited.

With the advent of molecular and genomic data and complementary analytical tools, the possibility of going beyond the traditional index gene finding to a more complex investigation of the genomic architecture of the placenta was deemed feasible.

Therefore, a comprehensive analysis of the placental gene expression in response to histamine was conducted with a combination of gene expression analysis technology including multiplex RT-qPCR and microarray technologies to provide better understanding of the effects of histamine in normal and pre-eclamptic placentae.

82 5.2 Histamine has regulatory feedback loop with Placental DAO (Publication #6 & #8)

A relationship between histamine concentration and diamine oxidase activity has been known for sometimes. It was reported as far back in the 60s that histamine in high concentration inhibits intestinal diamine oxidase activity in animal models

(Mondovi et al. 1965, Sasiak et al. 1975). Yet, this relationship was not tested in human placenta even though there was increasing evidence that maternal plasma diminished DAO activity concur with increasing maternal blood histamine (Section

2.2.4). Perhaps, this scarcity of evidence could or may have been partly due to lack of clarity on the source of increasing blood histamine and DAO in maternal blood. In an attempt to increase knowledge and our understanding of the relationship between histamine and DAO in the placenta, the effect of histamine on DAO mRNA expression was examined.

The study was guided by the hypothesis that increasing histamine concentration had inhibitory effect on placental DAO mRNA expression in placental explants. Therefore in a time series experiment, placental explants were treated with histamine at varying concentrations 0.01µM, 1µM and 100µM and cultured for 24 hours. Total RNA was extracted from time zero controls, time course histamine treated and non-treated

(negative controls) cultured placental explants and analysed with multiplex real time

RT-qPCR for DAO mRNA expression. Fold change expression relative to time zero samples was determined with ΔCT method using beta-actin as reference gene.

The findings showed that histamine regulated DAO mRNA expression, and the response was time and concentration dependent. Overall, 0.01µM histamine

83 concentration induced exponential increase in DAO expression with highest response occurring after 24 hours. The findings also showed that prolonged exposure to 1µM histamine concentration up-regulated DAO but not as strong as

0.01µM, whereas response to 100µM histamine was biphasic. At the high concentration (100µM), there was an immediate strong up-regulation followed by complete down-regulation after 24 hours.

Interestingly, the findings confirmed the previously identified (Mondovi et al. 1965,

Sasiak et al. 1975) dose dependent histamine inhibition of DAO activity and further showed that the effect could suggestively be intrinsically linked to histamine regulative effect on DAO gene expression in human placenta. Furthermore, the findings suggested that histamine appears to induce immediate concentration dependent increase in DAO mRNA expression, which if translated to protein would accelerate histamine metabolism. And while physiological levels of histamine

(0.01µM) increase placental DAO expression, long term exposure of placental explants to increasing histamine concentration appears to diminish DAO mRNA expression. These findings also showed a bi-directional feed-forward loop between histamine and DAO: two major biomarkers implicated in pre-eclampsia. In this loop, increasing histamine levels appear to trigger DAO mRNA expression which could lead to increased DAO protein activity and subsequent deamination and removal of accumulating histamine in the placenta. The findings also showed that excess production of histamine could down-regulate DAO mRNA expression leading to diminished DAO activity and further elevation of histamine levels in the placenta.

84 5.3 Histamine regulates placental Th-1/Th-2 regulation (Publication #6)

Human CD4+ and CD8+ T lymphocytes, macrophages, stromal cells express HDC,

H1R and H2R mRNA, and produce endogenously synthesised histamine (van der

Pouw Kraan et al. 1998, Hellstrand et al. 2000). Thus, changes in histamine concentration have been linked to activation of human T cells synthesis of Th-2 type cytokines, and down regulation of Th-1 cytokines (van der Pouw Kraan et al. 1998).

Typically, histamine through receptor H2 activation inhibits IL-12 mRNA synthesis, while up regulating IL-10 in monocytes or macrophages (van der Pouw Kraan et al.

1998, Packard & Khan 2003). Histamine also inhibits IL-2 and INF- synthesis in mice T lymphocytes (Kubo & Nakano 1999) and IL-1 production in mice macrophage

(Falus & Meretey 1992, Kubo & Nakano 1999). It is also shown that specific cytokines including IL-3 and GM-CSF regulate T cell production of histamine (Falus

& Meretey 1992, Kubo & Nakano 1999).

Relative synthesis of Th-1 and Th-2 type cytokines are implicated in the regulation of maternal immune response at the Foeto-Maternal Interface (FMI) (Wegmann et al.

1993). In this mechanism, it is suggested the foeto-placental unit redirects maternal immunity away from cell mediated immunity towards enhanced humoral responsiveness. In this case, it is observed that the balance between Th-1 and Th-2 cytokines is switched in favour of Th-2 cytokines (Wegmann et al. 1993) as activation of cell mediated immunity is usually harmful to the foetus (Rukavina & G

2000). CD4+ T helper Th-1 cells are known to secrete IL-2, IFN and TNF , where

IL-2 and TNF  in turn activates NK cells towards cell directed lysis (Rukavina &

Vince G, 2000; El et al., 2008; Laskarin et al., 2008). Both murine and human pregnancies with increased expression of Th-1 cytokines are highly unsuccessful

85 (Lin et al., 1993; Wegmann et al., 1993). Conversely, successful pregnancies tend to have increased expression of Th-2 cytokines (IL-4, IL-5, IL-10 and IL-13) (Rukavina

& Vince G, 2000).

Therefore, it was hypothesized that during normal pregnancy, histamine once produced could regulate Th- like cytokines in the placenta to modulate maternal immune response in favour of protecting the allogeneic foeto-placental allograft.

The effects of increasing histamine concentrations (10-7, 10-5, 10-3 M/L) respectively for low, medium and high dose on placental synthesis and production of cytokines were therefore evaluated. RNA was extracted from the control and histamine treated explants cultured for up to 24 hours and subsequently analysed with real time RT- qPCR. Gene expression was calculated with Pfaffl equation for the expression of

IL1beta, IL-10 and INF-γ, with beta-actin as housekeeping gene. The respective proteins were assayed with ELISA. Figure 5-1 A, C and E show the effects of histamine on IL1beta, IL-10 and INF-γ mRNA expression respectively. The corresponding protein levels are shown in Figure 5-1 B, D and F respectively.

The results showed that overall, histamine in low dose up-regulated (Figure 5-1) the expression of IL-10, IL-1β and INF- mRNA in term placenta after prolonged exposure, although there was an immediate inhibitory effect within 2 hours of IL-10, and INF- gene expressions. In contrast, high dose histamine inhibited the mRNA expression of IL-10 and INF- transcriptomic expression after prolonged exposure. In the case of IL-1β, the low dose histamine had minimal effect on the mRNA

86 expression after 24 hours but high concentration transiently increased the expression levels with peak effect occurring after 12 hours.

Figure 5-1: Effects of Histamine and Cytokine Synthesis and Production Figures A, C and E represent effects of histamine on IL1beta, IL-10 and INF-γ mRNA expression respectively. Figures B, D and F show the respective protein synthesis for IL1beta, IL-10 and INF-γ after 24 hours. Expression and protein levels are baseline (time zero) normalized.

The changes in the transcriptomic expression were reflected in the message translation into protein. As shown in Figure 5-1 B, D and F, the most persistent change in response to low dose histamine was observed after 24 hours of culture: low dose histamine caused gradual, yet sustained increase in IL-1beta, IL-10 and

87 INF-γ levels while high dose transiently increased the protein levels with peak effect occurring after 6 hours.

Also, while the individual gene expression data appeared to suggest that the effect of histamine on the respective genes after 24 hours could be minimal, analysis of the relative expression of INF- to IL-10 representing TH-1/Th-2 index (Figure 5-2) revealed a contrasting outcome. These findings (Figure 5-2) showed that the relative up-regulation of the Th-1 cytokine INF- in response to prolonged exposure of high dose histamine was greater (p< 0.001) than that of IL-10 (Th-2 type cytokine). In contrast, the relative expression of INF- to IL-10 mRNA expression at low dose histamine was significantly (p=0.018) decreased (i.e., Th-1/Th-2 ratio), therefore skewing the balance of the Th-1/Th-2 ratio in favour of Th-2 response in low dose histamine.

1.4

1.3

1.2

1.1 control a 10-7M 1 b 10-5M c 10-3M

.9 Ratio of INF-g/IL-10 Expression of Ratio .8

.6 2hr 6hr 18hr 24hr

Figure 5-2: Ratio of INF-γ and IL-10 Expression in the Placenta

88 These findings suggested that low dose histamine relatively up-regulated both Th-

1/Th-2 response in the placenta but favoured relative abundance of Th-2 response.

These findings also confirmed previous reports that physiological levels of histamine support humoral immune response in the placenta, in favour of Th-2 up-regulation towards positive pregnancy outcome (Falus & Meretey 1992, Wegmann et al. 1993, van der Pouw Kraan et al. 1998, Weetman 1999, Lim et al. 2000, Pap et al. 2007).

5.4 Uncovering the complexities of PE genetic pathways (Publication #2) Having established that there was a clear distinction between the levels of maternal blood histamine in PE and in normal pregnancy (Section 2.2.3); it was felt necessary to ascertain the extent to which placental gene expression differed between normal and pre-eclamptic placentae. A systematic review with meta-analysis of placental transcriptomic expression arrays was therefore conducted: first, to define the gene sets associated with normal and PE placentae; and secondly, to ascertain whether there was any relationship between the genes associated with PE and the clinical and or pathological presentation of the condition.

Using Absolute Gene Expression (AGE) methodology (section 3.4.5.1), significant genes defined as genes that were consistently up-regulated or down-regulated in PE and normal placentae (NP) were identified. Out of a total of 16,701, 5146 (31%) and

4394 (26%) placental genes were respectively identified as significantly down- regulated and up-regulated in the PE placentae (Table 5-1). In contrast, 846 and

1076 genes were respectively down-regulated and up-regulated in normal placentae

(Table 5-1). The expression levels of 14779 and 7161 genes respectively in NP and

89 PE placentae were inconsistent and thus were classified as non-significant (Table

5-1).

Table 5-1: Differentially Expressed Genes in NP and PE Placentae

Absolute PE only Absolute NP only Negatively Significant Genes # of down-regulated Significant Genes 5146 846 % of down-regulated Significant Genes 31% 5% Positively Significant Genes # of up-regulated Significant Genes 4394 1076 % of up-regulated Significant Genes 26% 6% Non-Significant Genes # of Non-Significant Genes 7161 14779 % of Non-Significant Genes 43% 88% A total of 167 microarray samples (PE = 68; NP = 99) were meta-analysed as case-to-control matched samples with Fisher’s method or as case (PE) only and control (NP) only with RankProd analysis. About 31% more genes were identified as differentially expressed (DE) in PE only than in case matched baseline (NP) subtraction PE. RankProd analysis Confidence at (1 - alpha): 95.0%; False Significant Proportion: 0.05 or less; p value threshold for fisher’s metaP = 0.05. PE = Pre- eclampsia Placentae; NP = Normal Placentae.

Bioinformatics analyses of the significantly regulated genes showed that the genes were involved in 207 and 126 biological pathways respectively for PE and NP.

Interestingly, all the 126 pathways enriched by the NP placental genes were also enriched in PE, but with higher enrichment ratios. A further 81 biological pathways were significantly enriched by PE placental genes only. Crucially, a number of pathways that have been associated with clinical manifestation of PE, including allograft rejection, apoptosis, VEGF signalling pathway, notch signalling pathway, and p53 signalling pathway, were significantly enriched in PE only. In addition, pathways such as histidine metabolism pathway and Fc epsilon RI-mediated signalling pathway were also significantly enriched in PE placentae only. These pathways have known histamine mediated functions. Thus, suggesting that histamine has some functional roles in the placenta and could be involved in the pathogenesis of PE.

90 For example, Fc epsilon RI-mediated signalling pathway is a pathway in mast cells.

Mast cell concentration is significantly increased in PE placentae (Purcell 1992,

Szewczyk et al. 2005). Fc epsilon RI-mediated signalling in mast cell is initiated by interaction of multivalent allogens with the extracellular domain of the alpha chain of

Fc epsilon RI to release pre-formed histamines, proteoglycans (especially heparin), phospholipase A2 and subsequently, leukotrienes (LTC4, LTD4 and LTE4), prostaglandins (especially PDG2), and cytokines including TNF-alpha, IL-4 and IL-5

(Kambayashi & Koretzky 2007). These mast cell activated mediators and cytokines levels are increased in PE, and increasing levels of these products have been associated with inflammatory responses.

Inflammation pathways have been strongly linked to the pathophysiology of PE

(Ilekis et al. 2016), and it has been suggested that the nuclear factor kappa-light- chain-enhancer of activated B cells (NF-κB) pathway mediates excessive maternal intravascular inflammation that leads to endothelial dysfunction (Aban et al. 2004,

Redman et al. 1999). In this context, it has been hypothesised that PE arises as a result of an excessive maternal intravascular inflammatory response to pregnancy, and that it involves the activation of both innate and the adaptive immune system, neutrophil, and the complement system pathways which involve both histamine and mast cell signalling (Greer et al. 1991, Falus & Meretey 1992, Prieto et al. 1997,

Redman et al. 1999, Belo et al. 2003, Lynch et al. 2008).

Similarly, allograft rejection, graft vs. host disease, and primary immunodeficiency pathways were significantly enriched in PE. This observation is consistent with

91 previous opinions that heightened immune responses in PE pregnancies could be a consequence of chronic foeto-allograft rejection reaction (Martinez-Varea et al.

2014). Accordingly, PE shares similarities with graft rejection linked to over activation of immune pathways (Wegmann et al. 1993, Weetman 1999, Raghupathy 2001,

Saito & Sakai 2003, Dong et al. 2005, Wilczynski 2006). Integral to this is the argument that disequilibrium of Th-1/Th-2 cytokine balance in favour of Th-1 (IL-2,

IL-12, IL-15, IL-18, IFNgamma, TNFalpha vs. IL-4, IL-10, TGFbeta); precipitation of subsets of immunocompetent cells (T CD4, suppressor gammadeltaT, cytotoxic T

CD8, Treg, Tr1, uterine NK cells); innate immunity (NK cytotoxic cells, macrophages, neutrophils and complement); adhesion molecules; fgl2 prothrombinase activation

(Wegmann et al. 1993, Weetman 1999, Raghupathy 2001, Saito & Sakai 2003,

Dong et al. 2005, Wilczynski 2006) and under-expression of Heme oxygenase-1

(HO-1) (Linzke et al. 2014) underpin the development of PE.

These findings not only confirmed the previous reports that inflammation and immunologic pathways are involved in PE pathophysiology, but also made major contributions to the knowledge and understanding of PE by providing evidence for a comprehensive database of gene sets with impacts on biomarker synthesis and early screening for PE. The PE immuno-inflammatory paradigm is however not universally supported as a recent review by Ahmed and Ramma (Ahmed & Ramma

2015) appears to down-play the roles of inflammatory, hypoxia and immunologic pathways in favour of angiogenic response as the cause of PE.

92 Ahmed and Ramma (Ahmed & Ramma 2015) argued that recent work supports the hypothesis that PE arises because of the loss of vascular endothelial growth factor

(VEGF) activity, which in turn is caused by increase in the levels of endogenous soluble fms-like tyrosine kinase-1 (sFlt-1), an anti-angiogenic factor (Ahmed &

Ramma 2015). SFlt-1 binds and reduces free circulating levels of the pro-angiogenic factor VEGF, and thus inhibits the beneficial effects mediated by flt-1 (also known as vascular endothelial growth factor receptor 1 (VEGFR-1)) on maternal endothelium, with consequential maternal hypertension and proteinuria (Andraweera et al. 2012,

Ahmed et al. 1997).

It is further argued that altered balance of circulating pro-angiogenic/anti-angiogenic factors such sFlt-1, soluble endoglin, and placenta growth factor (PlGF) are unique to PE (Andraweera et al. 2012, Ahmed et al. 1995, Ramma et al. 2012, Ahmed &

Ramma 2015, Ahmed et al. 1997). Interestingly, we also found evidence to support this view: for it was identified that VEGF signalling pathway was significantly enriched only by the PE placental genes. This therefore led to a conclusion that biological pathways affected in PE are complex, and instead of the uni-lateral conclusion drawn by Ahmed and Ramma (Ahmed & Ramma 2015), our findings offered more evidence to support the more global view that multiple and concurrent dysregulated pathways underpin the aetiology of PE (Ilekis et al. 2016), and no single pathway could be associated with the origins of PE. Thus, this work made major contributions to the knowledge and understanding of the molecular pathology or the genomic basis of PE, and provided further evidence for the links between histamine and PE.

93 5.5 Histamine regulates specific genes in pre-eclamptic placenta (Publications #9, 10 & #11) While it is widely accepted that the placenta is a major source of elevated histamine in the blood of women with PE (section 2.2.2), the placenta concurrently releases highly active diamine amine oxidase (DAO) enzymes into the maternal circulation to metabolise the histamine, thus forming histamine-DAO-Axis (HDA) (section 2.2.3).

However, in contrast to the effective clearance of histamine in normal pregnancy, the histamine-DAO-axis in PE is defective (section 2.2.4). Yet, our understanding of the impact of the mechanistic failure of the HDA and the consequential effects of elevated histamine on placental gene expression in PE was tenuous.

Therefore, an ex vivo elevated histamine model (EHM) in human placental explants with some parallel characteristic to in vivo defective histamine – DAO – axis (dHDA) system was developed (section 3.2.5 above) and used to examine the effects of elevated histamine on the placental gene expression. The work was supported by the hypotheses that (1) genes regulated by histamine in elevated histamine placental model would show similar changes in expression to PE specific placental genes and

(2) genes regulated by elevated histamine in an EHM would have known functions that could be biologically relevant in PE placentae.

5.5.1 Global Changes in EHM Gene Expression First, the gene expression patterns among histamine treated (also known as EHM) and control explants were compared to determine if the two classes differ with regard to expression profiles and whether the differences were significant. The BRB

ArrayTools platform (Section 3.4.5) was used for the analyses. As reported in publications #3 and #4, quality control (QC) of the microarray raw data was

94 assessed with Expression Console (Affymetrix) for .CEL files integrity. Probes with unusual signal patterns or signal strength and arrays showing low correlation between hybridization controls thus failed this initial QC measures were excluded from further analysis. The .CEL raw data were imported and processed with Robust

Multi-array Average (RMA) into BRB-Array Tools version 4.5.1 – Stable (Simon et al.

2007), and further processed using the R bionconductor packages (including Affy, annotate, annaffy, gcrma, simpleaffy) into log2-transformed and quantile-normalised linear model expression matrix. Genes showing minimal variation across the set of arrays were excluded from the analysis. Genes whose expression differed by at least

1.5 fold from the median in at least 20% of the arrays were retained. After reducing multiple probeset to singles using the most variable probeset measured by IQR across arrays, 20233 genes were used for further analyses.

The changes in gene expression in response to histamine treatment relative to the control under Physiologic oxygen concentration (8% O2) were established. Increases were determined for mRNA that were considered present or absent in the control and present in the histamine treated sample. Similarly, decreases were determined for mRNA that were present in the control sample and present or absent in the histamine treated sample. Scatterplot of Phenotype Averages (SPA) and Class

Comparison analyses (CCA) (Simon et al. 2007) were used to establish the difference in gene expression.

For the SPA analysis, average log-ratios within one phenotype (control samples), were plotted on the x-axis against the average log-ratio within the other phenotype

(histamine treated samples), on the y-axis in a scatterplot. These averages were then taken on a gene-by-gene basis, and each gene was represented by a single point in the resulting scatterplot. The differentially expressed genes among the two

95 phenotypes were determined as genes whose fold-change between the geometric mean of the expression ratios within each of the two classes was greater than a fixed threshold (1.5-fold change). Fold-change was defined as the ratio of geometric mean of intensities in class 1 to geometric mean of intensities in class 2, where class 1 was the treatment group and class 2, the control group. Thus, fold-change of less than 1 was regarded as down-regulation, and that greater than 1 as up-regulated. The SPA analysis identified (Figure 5-3A) a total of 494 differentially expressed genes (315 up-regulated and 179 down-regulated) between the EHM and Control samples.

Figure 5-3: Plots of differentially expressed among EHM and Control Samples Figure A shows the Scatterplot of differentially expressed genes with 1.5-fold change among elevated histamine model (EHM) control samples. Figure B shows a volcano plot of significant (p<0.05) differentially genes among EHM and control classes, identified with Class Comparison analysis. EHM = histamine treated samples. Appendix 3, 4 & 5 contain gene lists for both figures A and B. Expression patterns of the genes were confirmed by performing RT-qPCR on three randomly selected genes (AKR1C1, IGFBP5, and EREG relative to L19; Appendix 9).

To determine whether the differences in expression observed among the EHM

(histamine treated) and control classes were significant, class comparisons analysis using a random-variance t-test was performed (Figure 5-3B). Figure 5-3B shows the

96 results of the Class comparison analysis as a Volcano Plot of the significantly expressed genes among EHM and the controls.

In the Class comparison analysis, the purpose was to establish a general trend in the class differences with regard to expression profiles therefore a modest p value of less the 0.05 was deemed sufficient (Simon et al. 2007). The Class Comparison analysis was used because it is a powerful model that permits as was intended, the sharing of information among genes about within-class variation without assuming that all genes have the same variance (Wright & Simon, 2003).

Three key facts were learnt from these analyses:

1. That elevated histamine alters placental gene expression

2. That the relative differences in expression of specific genes between elevated

histamine model and control sample are significant.

3. While the analyses identified specific genes with significant differential

expression between the EHM and control samples, the information revealed

very little insight into the plausible roles or effects of the elevated histamine

regulated significant genes in the pathophysiology of PE placentae.

Therefore further innovative bioinformatics analyses were undertaken to decipher the plausible effects of elevated histamine on the placenta in the pathophysiology of PE

5.5.2 Determination of the plausible effects of histamine in PE Placentae In the preceding section, the point was raised that the mere comparison of differential gene expression between EHM and control samples provided limited insight into the plausible role of placental histamine in PE pathophysiology. This view was drawn from the fact that in addition to PE, defective histamine – DAO – Axis

97 leading to elevated histamine occurs in other trophoblast related complications of human pregnancy such as pre-term labour and spontaneous abortion. Therefore, in order to determine the effects or roles of elevated histamine in PE placentae, innovative yet robust down-stream analyses were performed. The purpose of these down-stream analyses was to identify specific elevated histamine regulated placental genes relevant to PE.

To achieve this, intra-class absolute gene expression analysis (Section 3.4.5.1) and

In Vivo Validation of In vitro Expression protocols were developed. PE placental genes were identified from the pool of genes regulated in EHM. The In Vivo validation of In vitro Expression is a complex bioinformatics workflow involving a systematic review process, meta-analysis, comparison analysis and confusion matrix analysis.

First, GEO and ArrayExpress databases were systematically searched for case control pre-eclampsia OR histamine AND Placenta Homo sapiens RNA array assays. Multiple RNA array assays for PE matched with NP placentae were retrieved. Intra-class Absolute Gene Expression analysis using RankProd statistics was applied to identify significant genes in EHM model, PE placentae and normal placentae respectively. The PE genes were defined as a set of specific genes that are consistently expressed exclusively in PE placentae but not in normal placenta. A variant definition for PE genes in EHM model was given as the set of consistently expressed genes in response to elevated histamine in EHM that mapped to genes consistently expressed in PE placentae but not to consistent significantly expressed genes in normal placentae, or to cultured control sample.

98 The analyses identified 270 genes that were consistently expressed in the elevated histamine model and mapped to PE placental specific genes. That is, these genes were not consistently expressed in normal pregnancy placentae or cultured explants without histamine treatment, and thus was classified as histamine regulated specific genes in PE (HSPE). The overall concordance between the expression pattern of the

HSPE genes and the PE specific genes, measured as accuracy and intra-class correlation coefficient were 92%, Kappa = 0.836 respectively. Also, 94.3% of the down-regulated HSPE genes were down-regulated in PE placentae. Similarly, 90% of the up-regulated HSPE genes were up-regulated in PE placentae. These suggested that the pattern of expression of the HSPE genes closely mimicked gene expression pattern of the in vivo PE placentae.

With this clarity, the biological functions of the HSPE genes were examined with in- depth bioinformatics pipeline including Over-representation of Gene Ontologies,

Pareto Analysis for vital few genes and GSEA Leading Edge Metagene (LEM) analyses of biological pathways for leading edge genes. The analyses uncovered complex gene ontologies (Table 5-2) that offered stronger elucidation of plausible histamine functions in PE placentae. Analysis for the vital few genes revealed that 51

(vital few) of the 270 HSPE genes (Table 5-3) were involved in 80% of the associated biological functions. Downstream analysis with leading edge meta-genes analysis further identified 24 enriched pathways (Table 5-4) from 4 databases

(KEGG, Reactome, Wikipathways and Panther) that histamine specific consistent regulated genes could play significant roles.

99 Table 5-2: Histamine Specific Genes in PE Ontologies GO GO Term ID Name C O E R Pvalue Category MF GO:0002474 antigen processing and 87 9 1.17 7.71 2.46E-06 presentation of peptide antigen via MHC class I MF GO:0002064 epithelial cell development 195 12 2.61 4.59 1.30E-05 MF GO:0060684 epithelial-mesenchymal cell 7 3 0.09 31.96 8.00E-05 signaling MF GO:0060710 chorio-allantoic fusion 7 3 0.09 31.96 8.00E-05 MF GO:0048002 antigen processing and 173 10 2.32 4.31 1.16E-04 presentation of peptide antigen MF GO:0006397 mRNA processing 463 17 6.21 2.74 1.75E-04 MF GO:0003334 keratinocyte development 11 3 0.15 20.34 3.62E-04 MF GO:0010608 posttranscriptional regulation 454 16 6.09 2.63 4.26E-04 of gene expression MF GO:0001738 morphogenesis of a polarized 134 8 1.8 4.45 4.51E-04 epithelium MF GO:0071456 cellular response to hypoxia 136 8 1.82 4.39 4.98E-04 CC GO:0005913 cell -cell adherens junction 318 16 2.9 5.51 3.65E-08 CC GO:0015629 actin cytoskeleton 456 16 4.16 3.84 4.46E-06 CC GO:0044297 cell body 472 16 4.31 3.71 6.89E-06 CC GO:0005925 focal adhesion 390 14 3.56 3.93 1.41E-05 CC GO:0005924 cell -substrate adherens 393 14 3.59 3.9 1.53E-05 junction CC GO:0030055 cell -substrate junction 398 14 3.63 3.85 1.76E-05 CC GO:0043292 contractile fiber 209 10 1.91 5.24 2.26E-05 CC GO:0098858 actin -based cell projection 180 9 1.64 5.47 4.16E-05 CC GO:0043025 neuronal cell body 411 13 3.75 3.46 1.05E-04 CC GO:0005681 spliceosomal complex 173 8 1.58 5.06 1.92E-04 MF GO:0098641 cadherin binding involved in 277 14 3.24 4.32 4.85E-06 cell-cell adhesion MF GO:0098632 protein binding involved in cell- 288 14 3.37 4.16 7.58E-06 cell adhesion MF GO:0098631 protein binding involved in cell 293 14 3.42 4.09 9.22E-06 adhesion MF GO:0045296 cadherin binding 294 14 3.44 4.07 9.59E-06 MF GO:0045505 dynein intermediate chain 7 3 0.08 36.66 5.32E-05 binding MF GO:0003779 actin binding 390 15 4.56 3.29 5.60E-05 MF GO:0050839 cell adhesion molecule binding 445 16 5.2 3.08 7.02E-05 MF GO:0005520 insulin -like growth factor 28 4 0.33 12.22 2.97E-04 binding MF GO:0031994 insulin -like growth factor I 12 3 0.14 21.39 3.20E-04 binding MF GO:0010385 double -stranded methylated 5 2 0.06 34.22 1.33E-03 DNA binding C: the number of reference genes in the category; O: the number of genes in the user gene list and also in the category; E: The expected number in the category; R: ratio of enrichment; PValue: p value from hyergeometric test.

100 Table 5-3: Histamine Regulated Genes Performing 80% of Histamine Mediated Functions in PE Rank GeneSymbol Name Count of Sign Functions 1 FLNB filamin B, beta (actin binding protein 278) 14 (+ve) 2 ITGA6* integrin, alpha 6 12 (+ve) 3 HSPA1A heat shock 70kDa protein 1A 11 (+ve) 4 MPRIP myosin phosphatase Rho interacting protein 11 (+ve) 5 CDC42 cell division cycle 42 (GTP binding protein, 25kDa) 10 (+ve) 6 PHLDB2 pleckstrin homology-like domain, family B, member 2 9 (+ve) 7 CALD1 caldesmon 1 9 (+ve) 8 PICALM phosphatidylinositol binding clathrin assembly protein 8 (+ve) 9 RAN RAN, member RAS oncogene family 8 (+ve) 10 PALLD palladin, cytoskeletal associated protein 8 (+ve) 11 TJP1 tight junction protein 1 (zona occludens 1) 7 (+ve) 12 RAB1A RAB1A, member RAS oncogene family 7 (+ve) 13 LARP1 La ribonucleoprotein domain family, member 1 7 (+ve) 14 KIF5B kinesin family member 5B 6 (+ve) 15 GJA1 gap junction protein, alpha 1, 43kDa 6 (+ve) 16 TPM4 tropomyosin 4 6 (+ve) 17 PAFAH1B1 platelet-activating factor acetylhydrolase, isoform Ib, alpha subunit 6 (+ve) 45kDa 18 BZW1* basic leucine zipper and W2 domains 1 6 (+ve) 19 OPRM1 opioid receptor, mu 1 5 (-ve) 20 ACTR3 ARP3 actin-related protein 3 homolog (yeast) 5 (+ve) 21 PDIA3* protein disulfide isomerase family A, member 3 5 (+ve) 22 SEPT2 septin 2 5 (+ve) 23 ARF1 ADP-ribosylation factor 1 5 (+ve) 24 PSMD14 proteasome (prosome, macropain) 26S subunit, non-ATPase, 14 4 (+ve) 25 PSMA2 proteasome (prosome, macropain) subunit, alpha type, 2 4 (+ve) 26 PSMD4 proteasome (prosome, macropain) 26S subunit, non-ATPase, 4 4 (+ve) 27 PSME4 proteasome (prosome, macropain) activator subunit 4 4 (+ve) 28 ZFP36L1 zinc finger protein 36, C3H type-like 1 4 (+ve) 29 ABRA actin-binding Rho activating protein 3 (-ve) 30 RALA v-ral simian leukemia viral oncogene homolog A (ras related) 3 (+ve) 31 APP amyloid beta (A4) precursor protein 3 (+ve) 32 IQGAP2** IQ motif containing GTPase activating protein 2 3 (-ve) 33 TPM2 tropomyosin 2 (beta) 3 (+ve) 34 ADAR adenosine deaminase, RNA-specific 3 (+ve) 35 FOXA1 forkhead box A1 3 (-ve) 36 TRPC5 transient receptor potential cation channel, subfamily C, member 5 3 (-ve) 37 GPM6A glycoprotein M6A 3 (-ve) 38 MPL myeloproliferative leukemia virus oncogene 3 (-ve) 39 CYR61 cysteine-rich, angiogenic inducer, 61 3 (+ve) 40 SYNCRIP synaptotagmin binding, cytoplasmic RNA interacting protein 3 (+ve) 41 MTPN myotrophin 2 (+ve) 42 ANXA3 annexin A3 2 (+ve) 43 HNRNPH3 heterogeneous nuclear ribonucleoprotein H3 (2H9) 2 (+ve) 44 MYO3B myosin IIIB 2 (-ve)

101 45 GLRA3 glycine receptor, alpha 3 2 (-ve) 46 ARPC4 actin related protein 2/3 complex, subunit 4, 20kDa 2 (+ve) 47 SEC13 SEC13 homolog (S. cerevisiae) 2 (+ve) 48 DLG5 discs, large homolog 5 (Drosophila) 2 (+ve) 49 SLC5A7 solute carrier family 5 (choline transporter), member 7 2 (-ve) 50 DNAJB6 DnaJ (Hsp40) homolog, subfamily B, member 6 2 (+ve) 51 KHDRBS1 KH domain containing, RNA binding, signal transduction associated 1 2 (+ve) Count = number of GO terms gene is involved; (+ve) = consistently expressed at High levels; (-ve) = consistently expressed at low levels. Levels of expression of all genes matched the expression levels in PE placentae except where indicated: * signifies genes that are expressed at high levels in EHM but at low levels in PE placentae; ** genes expressed at low levels in EHM but at high levels in PE placentae.

Table 5-4: Histamine Mediated Pathways with Leading Edge Genes ID Name Size L ES NES WP455 GPCRs, Class A Rhodopsin-like 8 8 -0.62 -1.71 hsa04080 Neuroactive ligand-receptor interaction - Homo 8 8 -0.62 -1.84 sapiens (human) hsa05200 Pathways in cancer - Homo sapiens (human) 9 8 0.61 1.4 R-HSA-500792 GPCR ligand binding 10 8 -0.32 -0.97 WP2431 Spinal Cord Injury 7 6 0.43 0.95 hsa04144 Endocytosis - Homo sapiens (human) 6 6 0.69 1.47 WP2377 Integrated Pancreatic Cancer Pathway 6 5 0.56 1.18 WP411 mRNA Processing 5 5 0.48 0.97 hsa04260 Cardiac muscle contraction - Homo sapiens 5 5 0.75 1.54 (human) hsa04932 Non-alcoholic fatty liver disease (NAFLD) - Homo 5 5 0.81 1.64 sapiens (human) hsa04072 Phospholipase D signaling pathway - Homo 5 5 0.7 1.41 sapiens (human) hsa04015 Rap1 signaling pathway - Homo sapiens (human) 6 5 0.64 1.35 hsa05132 Salmonella infection - Homo sapiens (human) 5 5 0.68 1.37 P00034 Integrin signalling pathway 6 5 0.56 1.19 R-HSA-425407 SLC-mediated transmembrane transport 5 5 -0.62 -1.42 R-HSA-418555 G alpha (s) signalling events 5 5 -0.56 -1.28 WP138 Androgen receptor signaling pathway 5 4 0.71 1.46 WP481 Insulin Signaling 6 4 0.69 1.48 WP306 Focal Adhesion 5 4 0.43 0.86 hsa05010 Alzheimer's disease - Homo sapiens (human) 6 4 0.73 1.57 hsa04022 cGMP-PKG signaling pathway - Homo sapiens 5 4 0.78 1.59 (human) hsa04010 MAPK signaling pathway - Homo sapiens 5 4 0.64 1.29 (human) P00016 Cytoskeletal regulation by Rho GTPase 5 4 0.43 0.89 WP2380 Brain-Derived Neurotrophic Factor (BDNF) 5 2 0.4 0.81 signaling pathway Table is sorted according to L. Size: the number of genes in the user gene list and also in the category; L: the number of leading edge genes; ES: Enrichment Score; NES: Normalized Enrichment Score; (-) = down-regulated pathway

102

This work showed a strong correlation between significantly enriched GOs (Table

5-2) and the pathways identified with leading edge meta-gene analysis (Table 5-4).

For example, pathways including Insulin Signaling, Focal Adhesion, mRNA

Processing, Salmonella infection and muscle contraction are very closely related to

GO terms such as insulin-like growth factor I binding, focal adhesion, mRNA processing, antigen processing and presentation of peptide antigen, dynein intermediate chain binding, and actin binding. Thus, confirming that genes consistently regulated by elevated histamine in an EHM model had functional roles that were biologically relevant in PE placentae.

5.5.3 Identification of causal effect of histamine specific genes in PE In order to determine plausible causal effect of the histamine regulated genes in PE,

Directed acyclic graphs (DAGs) causal analysis was performed using dagitty.net

(Textor et al. 2011). This study was aimed at (1) identifying the minimal sufficient adjustment sets of information from the GOs and biological pathways enriched by histamine regulated genes in PE; (2) to identify the Instruments and conditional variables, the adjustments necessary to estimate the total effect, and the minimal sufficient adjustment sets for estimating the direct effect of elevated histamine on pre-eclampsia to derive testable implications.

To achieve these, extensive literature review using thematic analysis was firstly conducted on the functions of the genes enriching the significant GOs and pathways to document the covariates’ associations with the elevated histamine (exposure) and

PE (outcome). DAG tool was then used to identify the minimal adjustment set of covariate to minimize the magnitude of bias (Figure 5-4A).

103

Figure 5-4: Causal effect identification of Elevated Histamine on Pre-eclampsia DAG Direct Acrylic Graph for the 9 major functional groupings for histamine regulated genes in PE (Figure A). Figure B shows the DAG causal pathways between histamine and PE and regulated genes functions covariates. exposure; outcome; ancestor of exposure; ancestor of outcome; causal path

The analysis enabled the identification of 9 minimal sufficient adjustment functional group covariates (Figure 5-4A) with 54 sub-functional group covariates and 45

104 causal paths (Figure 5-4B) to estimate direct effect of ‘Elevated Histamine’ on Pre- eclampsia. The analysis also revealed that no adjustment to the minimum functional groups was necessary to estimate the total effect. Nonetheless, it clarified the roles of DAO and HDC in PE by predicting that defective DAO and increased HDC activities are instrumental and conditional variables for the estimating the causal effect of elevated histamine in PE.

Also, the analysis clarified that the likelihood that elevated histamine will directly cause PE is nominal but greatly through the regulation of specific genes that activate minimal sufficient functional covariate and sub-functional groups. The minimal but major covariate functional groups included: Cell Signalling, Gene Regulation,

Inflammation and Immune regulation, Junctional proteins regulation, Oxygen

Sensing, Placental Innervation, Placental contractile activity, Placental growth,

Tissue proliferation and morphogenesis.

This study was very important to furthering our understanding of the relationship between histamine and PE. In that, in addition to generating 57 testable implications for future work, it refined our knowledge that (1) the natural process which determines the amount of histamine in the placental and maternal blood is affected by DAO and HDC activity; (2) the natural process by which pre-eclampsia develops is affected by the amount of histamine in the maternal blood and placenta; (3) the natural process by which pre-eclampsia develops is not affected by the DAO and

HDC activities other than indirectly via the histamine deposit; and finally (4) the natural process by which pre-eclampsia develops is indirectly affected by histamine via functional group covariates, and causal paths.

105 5.6 Discussions and Conclusion Data mining for the functions and relevance of the enriched biological pathways revealed a complex network of functions, yet a better understanding of the effects of elevated histamine in the placenta. While histamine is generally known to regulate several physiological processes, including neurotransmission, gastric acid secretion, inflammation, and smooth muscle tone, elevated histamine in the placenta through this cluster of studies have been shown to regulate the expression of vital genes with complex functions that are implicated in PE.

The study examining the relationship between histamine and DAO was the first to demonstrate that these biomarkers implicated in pre-eclampsia form a bi-directional regulatory negative feed-forward relationship in the placenta. The causal effect analysis clarified that the decreased activity of DAO and increased HDC activity are actually instrumental and conditional variables that are expected for elevated histamine to occur in PE placentae. Using a self-validating gene pool, our understanding of the effect of histamine in the pathophysiology of PE was further broadened. The study of the effect of elevated histamine on placental explants was the first to successfully model ex vivo defective histamine-DAO-axis to provide an in vitro culture model to investigate the effect of elevated histamine in parallel with in vivo PE placentae. From these works, it was also shown for the first time that PE placentae express unique set of significant genes in addition to the significant genes expressed in NP, and that elevated histamine regulated a subset of these PE specific genes in vitro.

106 This revelation provided further unique insight into the biological pathways through which elevated histamine would seem to exert effects in the pathophysiology of PE.

For example, the identification of the functional group covariate ‘Placental

Innervation’ consisted of leading edge genes that formed the sub-functional covariates including Neuroactive ligand-receptor interaction - Homo sapiens

(human), Brain-Derived Neurotrophic Factor (BDNF) signalling pathway and Spinal

Cord Injury regulating. The Placental contractile activity functional covariate group consisted of sub-groups including: muscle contraction including Cardiac muscle contraction - Homo sapiens (human) pathway, actin cytoskeleton, contractile fiber, actin-based cell projection, dynein intermediate chain binding, actin binding and

Cytoskeletal regulation by Rho GTPase pathways.

These findings were fascinating, nonetheless unexpected because the placenta is currently recognised as not an innervated organ (Walker & McLean 1971, Khong et al. 1997, Marzioni et al. 2004). The findings therefore provided new opportunity to re- evaluate this long-held view that the placenta is de-innervated. The outcome of this re-evaluation was that currently, based on limited publications there is no evidence that the placenta is innervated. However, our work provided fresh evidence to justify further studies to examine whether elevated histamine levels in placentae destined to develop PE could regulate subsets of genes that might preserve innervation and contractility associated with PE placental vasculature (Reilly & Russell 1977,

Szukiewicz et al. 1999, Pijnenborg et al. 2006b). Alternatively, the findings offered the opportunity to ask whether there is yet an un-identified pathway in PE placentae that mimics Neuroactive ligand-receptor interaction - Homo sapiens (human), Brain-

Derived Neurotrophic Factor (BDNF) signaling pathway and Spinal Cord Injury

107 smooth muscle contraction pathways in the placenta, and thus warrant further investigation.

In addition to the generation of new theories about the functions of the placenta, the studies confirmed previous observations of the effects of histamine. It was previously reported that increasing histamine causes placental microvascular endothelial damage through redistribution of PECAM-1 and VE-cadherin molecules to non- junctional regions, thus inducing vascular permeability and exposure of the basement membrane and vascular muscles to vasoactive ligands (Leach et al. 1995,

Leach & Firth 1997, Chambers et al. 2001). It is also established that endothelial cell activation and dysfunction are also strongly associated with PE (Roberts et al. 1989,

Taylor et al. 1998). Our work not only confirmed these previous reports but provided further insight into the complex network of genetic pathways including cell-cell adherens junction, focal adhesion, cadherin binding involved in cell-cell adhesion, and protein binding involved in cell-cell adhesion through which elevated histamine could function to precipitate endothelial dysfunction in PE.

Notwithstanding the effect of histamine on inflammatory pathways in the placenta, the studies on the relationship between histamine and index cytokines provided both quantitative RNA and protein assays to confirm that histamine regulate pro- inflammatory cytokines including IL-1b, IL-10 and INF- in the placenta, and elevated histamine tended to favour PE linked Th-1 cytokine INF- response.

Taken together, the studies not only elaborated on the complexity of histamine’s function in the placenta, but added to our knowledge and understanding of the

108 pathogenesis of PE by showing that elevated histamine could significantly precipitate

PE through the regulation of key placental genes involved in inflammation and immune system regulation; tissue proliferation and morphogenesis; regulation of junctional proteins; maintenance of placental contractile activity; regulation of growth; impairment of cell signalling; and regulation of placental gene expression. Thus, confirming that histamine has significant mediating and or sustaining functions in the placenta, and that further studies to confirm causal effect of elevated histamine in PE pathophysiology is warranted.

109 Chapter 6 Final Conclusion

6.1 Reflection on contributions made by the publications to the knowledge and Science of Placental Genomics and Pre-eclampsia

In this commentary, I have critically reflected on the rationale for investigating the functional roles of histamine in human placenta with a focus on pre-eclampsia. I have also critically reflected on the suitability of the methods, results and the relative contribution made by the series of publications to the knowledge and understanding of placental genomics with reference to the effects of histamine in human placenta.

6.1.1 The placenta expresses histamine receptors, and histamine production and elimination in the PE placentae are impaired Prior to the onset of my research, it was known that maternal blood histamine levels are elevated during PE (Figure 2-4). However, due to lack of evidence for histamine functionality in the placenta, very limited progress was made to advance our understanding on the causal effects of the amine on PE. Most importantly, knowledge about the topological distribution of histamine receptor expression in human placenta was extremely limited. Nonetheless, through the preceding discussions, I have shown that one of my publications (Section 4.4) was the first to report that histamine receptors H1R and H2R are expressed in juxtaposed positions with DAO message at the foeto-maternal interface in the villous parenchyma.

Prior to this discovery, it was known that histamine receptors H1R and H2R have both constitutive and inductive activities in other systems, and the constitutive activity was associated with a basal whole blood histamine concentration within the 0.1M

110 range (plasma equivalent in 2.0 nM range) (Kahlson et al., 1960b; Kahlson &

Rosengren, 1971; Beaven, 1982; Bakker et al., 2000; Alewijnse et al., 2000). This view was postulated as a physiological switch for fine tuning specific cellular activities including tissue remodelling, immuno-modulation, cell proliferation and differentiation, tissue growth and wound healing (Kahlson, 1960; Kahlson et al.,

1960a; Kahlson & Rosengren, 1968; Kahlson & Rosengren, 1971; Kahlson &

Rosengren, 1972; Beaven, 1978; Erlik et al., 1979; Krishna et al., 1986; Krishna et al., 1989; Tetlow & Woolley, 2003). Nonetheless, the extent to which histamine modulates human placental tissue remodelling was unclear. Therefore, I undertook subsequent studies to generate new knowledge to enhance our understanding on the effects of histamine in human placenta and PE.

Importantly, with meticulous examination of the literature, I provided coherent knowledge and additional insight into the understanding of the profile and dynamics of maternal blood histamine levels in both normal and pre-eclamptic pregnancies

(section 2.2). I thus contributed to further our understanding of the nature of maternal blood histamine level profiling that while maternal blood histamine levels in the course of normal pregnancy generally fall below pre-pregnancy levels (Figure 2-2), the levels remain elevated in PE (Figure 2-4). I have also contributed to clarifying that the elevated histamine levels in PE are due to a defective histamine-DAO-axis

(section 2.2.4), which in turn could be caused by increased histamine synthesis

(section 4.2) and diminished DAO activity (section 5.2).

I provided specific evidence to contribute to knowledge on this matter by examining factors that could affect histamine production and elimination in the placenta, and

111 thus impact on PE. I showed that placental histidine decarboxylase activity is increased in PE placentae, and cytokines and mitogens increase histamine production in the placenta. I also showed that histamine has a feedforward regulatory loop with DAO, where increasing histamine levels cause down-regulation of DAO gene expression.

6.1.2 Elevated histamine in the placenta has functional roles: it regulates production of Th-1/Th-2 like cytokines in the placenta The immune-modulative effect of histamine was previously demonstrated in other systems to be dependent on the concentration of bioactive histamine. For example, increasing histamine concentration inhibited macrophage ROS production and thus reversed NK cells and CTL deactivation by macrophage (Mellqvist et al., 2000;

Hellstrand et al., 2000b). This was consistent with the clinical observations where decreased or falling DAO activity led to increased bioactive histamine and resultant spontaneous abortions (Southren et al. 1966b).

Therefore, I examined and generated further knowledge on placental histamine physiology that provided evidence to support the claims that histamine dose dependently activates Th-1/Th-2 response in the placenta (Figure 5-2). Here, I demonstrated that low dose histamine up-regulates Th-2-type cytokine responses, and that this fine tuning could sustain successful foetal allograft and pregnancy

(Wegmann et al., 1993) while high dose histamine up-regulated Th-1-type cytokine responses with potential to cause foetal demise (Weetman 1999). This contribution first and foremost, confirmed previous reports of Th-1/Th-2 switches in human placenta (Darmochwal-Kolarz et al. 1999, Saito et al. 1999, Saito & Sakai 2003). The

112 work further showed that the observed histamine dose dependent regulation of Th-

1/Th-2 response in peripheral tissues (Packard & Khan 2003) was also present in human placenta and that, varying concentrations of FMI circulating histamine does have key regulatory roles on the expression of the local Th-1/Th-2 cytokine genes.

6.1.3 Optimum culture condition is fundamental for appropriate investigation of histamine effect in human placenta and implications in PE Having established that the placenta expresses histamine receptors and histamine has bi-directional regulatory effect with DAO and key index cytokines (sections 4.3 and 5.3), further investigations involving genome wide transcriptomic expression analyses were used to study the effects of histamine on the global expression of placental gene with a focus on PE. In the first instance, an ex vivo defective

Histamine-DAO-axis also referred to elevated histamine model was developed

(section 3.2.5). In course of this investigation, I provided new evidence to confirm that matching in vitro oxygen tension with in vivo conditions (which vary with gestational age) was a key component for optimal condition for placental explant culture (section 3.2.4). And that oxygen levels could have marked effects on for example, trophoblast apoptosis during the placental explant cultures (Damsky et al.

1993, Genbacev et al. 1996, Caniggia et al. 2000, Burton & Caniggia 2001, Huppertz et al. 2003). I contributed further knowledge to provide clarity on the extent to which tissue culture per se, culture oxygen concentration and culture duration affect placental gene expression. I provided genetic evidence to confirm previous morphological observations (Burton & Caniggia 2001, Burton & Jauniaux 2004,

Burton et al. 2006) that high oxygen, i.e, atmospheric oxygen concentration (AOC) was not physiological for term placental culture. In this evidence, I showed that micro

113 explants were viable after 6 days culture in both AOC and Physiological Oxygen

Concentration (POC). However, there was marked difference in tissue morphology between explants cultured in POC and AOC. I provided new evidence that while the explants were viable after 6 days culture, there was more syncytial detachment and loss in explants cultured in AOC; and that the tissue morphology and RNA quality in explants cultured in POC was akin to pre-culture explants. I also showed for the first time that placental RNA quality decline in tandem with STB degeneration and syncytial damage and loss. Hence, I provided new knowledge to further our understanding that micro explants cultured at POC have the best mRNA quality and tissue structure after long-term culture to support high throughput experiments.

Regarding knowledge and understanding of the effects of placental explant culture on placental gene expression, I provided comprehensive new evidence to show that the observable morphological changes and differences in RNA quality between the

POC and AOC cultured explants transcended beyond morphology to deeper changes in the biological and molecular pathways in the placenta; and that the changes impacted on interpretations of studies that have used explants cultured in

AOC. I demonstrated further that explant culture in AOC not only affects individual gene expression but also induce relative changes to the enrichment of specific transcription factor targets, placental molecular functions, cellular components, biological processes and functional pathways.

114 6.1.1 Histamine has complex functions in the placenta: it regulates genetic pathways that have implications in PE placentae I then undertook further work to extend our knowledge and understanding of gene expression in normal and PE placentae (Section 5.4). This work was an extension to a previous report on PE gene expression meta-analysis, which explored the relative relationship between NP and PE gene expression. In this work, I broadened the coverage to examine the roles of genes with consistent low or high level expressions in the placenta. One key finding was the identification for the first time, of subsets of genes that are consistently up or down-regulated in PE but not in NP placentae. This enabled the identification of specific biochemical pathways that characterises the contributory roles or causal effects of elevated histamine in the pathogenesis of PE.

Through this work, I further showed that about 30% of genes currently published as being investigated for possible roles in the molecular pathology of PE were not consistently expressed in the PE placentae.

The work had interesting impact on further studies in PE. For, it not only provided a unique opportunity for researchers in the field to re-evaluate the relevance and focus of their investigations of PE, it offered further opportunities to refine the studies designed for molecular bio-markers for early detection of women at risk of preeclampsia. The work also developed on the framework for absolute gene expression analysis into modelling the baseline genomic blueprint for normal placental gene expression for comparison with expressions in complicated placentae.

Thus, this work was further extended to examine the effect of elevated histamine on placental gene expression in parallel with effect of PE on the placenta (Section 5.5).

115 In this extension work, new evidence was added to the knowledge and understanding of the roles of specific placental genes regulated by histamine and their implications in PE. The key finding was Histamine regulated a subset of placental genes that were significantly expressed in PE placentae but not in normal pregnancy placentae or cultured explants without histamine treatment. I provided evidence that the elevated histamine regulates specific genes in pre-eclampsia, and that the genes are involved in inflammation and immune system regulation; tissue proliferation and morphogenesis; regulation of junctional proteins; maintenance of placental contractile activity; regulation of growth; impairment of cell signalling; and regulation of placental gene expression. These suggested that histamine has significant mediating and or sustaining functions in the placenta, and that elevated levels do play key roles in the pathophysiology of pre-eclampsia.

6.1.2 Is histamine the missing link in PE pathophysiology? In the introduction, it was stated that if histamine receptors are expressed in the placenta, then elevated histamine would have functional roles, and if the effects of elevated histamine were similar to the pathophysiological changes observed in PE placentae, then histamine could be the missing link that precipitates or decompensate PE pathophysiology. In the course of these discussions, I have provided evidence through the series of publication submitted here for consideration that: histamine receptor genes are expressed at the foeto-maternal interface; DAO genes are expressed in the placenta; HDC activity levels are increased in PE placentae; there is cross-talk between histamine, DAO and cytokines in the placenta; elevated histamine in dHDA regulated the expression of specific genes in the placenta and these genes are abnormally expressed in PE placentae; the functions

116 of these elevated histamine regulated genes also expressed in PE are involved in tissue morphology and possibly poor placentation, metabolic defects, endothelia dysfunction, inflammation, immunologic response, angiogenic and anti-angiogenic response in PE placentae. It is therefore suffice to suggest that elevated histamine fits the role of the missing link (Figure 6-1) in PE pathophysiology and that elevated histamine has causal effects on PE and is plausibly, the missing link that perpetuates the pathophysiological changes in the placental and maternal systems to culminate in pre-eclampsia complication. Therefore it is reasonable to conclude that the elevated histamine observed in PE would have pathophysiological roles in PE and early detection leading to effective control of maternal blood histamine levels could save lives and it’s thus recommended.

117

Figure 6-1: Histamine fits the missing link in Pre-eclampsia Pathogenesis

118 6.2 Reflection on my contributions and my professional development as a research practitioner My contributions to the listed publications are described in Table 6-1 below according the CRediT Taxonomy of author contributions (Brand et al. 2015). My contributory role is described in column 1, followed by the role definition in column 2.

The respective publications are numbered from 1 – 11 in line with the publication portfolio listing. The listings and completion (√) of this information with copies of the published author contribution accurately reflect my contributions to the works submitted here for the award.

Table 6-1: Summary of My Contributory

My Role Definition 1 2 3 4 5 6 7 8 9 10 11 Contributory Roles Conceptualisation Ideas; formulation or √ √ √ √ √ √ √ √ √ √ √ evolution of overarching research goals and aims. Data Curation Management activities √* √* √* √* √* √* √* √* √* √* √* to annotate (produce metadata), scrub data and maintain research data for initial use and later reuse. Formal Analysis Application of statistical, √ √ √ √ √ √ √ √ √ √ √ mathematical, computational, or bioinformatics to analyse or synthesize study data. Investigation** Conducting research √ √ √* √ √* √ √ √ √ √ √ and investigation process, specifically performing the experiments Methodology Development or design √ √ √* √* √* √* √* √* √* √* √* of methodology; creation of models Project Management and √* √* √* √* √* √* √* √* √* √* √* Administration coordination responsibility for the research activity planning and execution. Resources Provision of study √* √* √* √* √* √* √* √* √* √* √* materials, reagents, materials, patients,

119 laboratory samples, animals, instrumentation, computing resources, or other analysis tools. Validation Verification, whether as √ √ √* √ √* √ √ √ √ √ √ a part of the activity or separate, of the overall replication/reproducibility of results/experiments and other research outputs. Visualization Preparation, creation √ √ √ √ √ √ √ √ √ √ √ and/or presentation of the published work, specifically visualization/data presentation. Writing – Original Creation and/or √ √ √ √ √ √ √ √ √ √ √ Preparation presentation of the published work, specifically writing the initial draft (including substantive interpretation). § § § § § § § § § § Review & Editing Presentation of the √* √ √ √ √ √ √ √ √ √ √ published work, specifically critical review, commentary or revision – including pre- or post-publication stages. √* = joint activity; √§ = I performed activity but consulted with named others; ** = list of specific experiments or investigations I performed (tissue dissection, tissue culture, protein and amine assays with ELISA, gene cloning, immuno-histochemistry; in situ hybridisation, RNA and DNA extraction, PCR, RT-PCR, real time RT-qPCR, systematic review, meta-analysis, all statistical analysis, all bioinformatics analysis)

Over the course of this work, it was a primary desire to improve on my skills for conceptualising research ideas. To achieve this, I used a 3Cs (Curiosity, Critical thinking, and Courage for idea generations) prong approach to improve on my conceptualisation skills. Development, harnessing and control of my curiosity for good research were particularly important for me. In the Old Man's Advice to Youth:

"Never Lose a Holy Curiosity," LIFE magazine (2 May 1955, p. 64) Einstein is famously quoted as saying: “The important thing is not to stop questioning. Curiosity has its own reason for existence. One cannot help but be in awe when he contemplates the mysteries… of life, of the marvellous structure of reality. It is

120 enough if one tries merely to comprehend a little of this mystery each day. Never lose a holy curiosity”. Thus, I desired to develop a higher sense of curiosity, a holy curiosity to continue to question, control my curiosity, amass courage to entertain and refine scientific questions including those others considered “bad ideas or red herrings” through critical thinking into clearly stated researchable ideas. The outcome for this is clearly outlined in Table 6-1, row 1. The ideas for all the studies underpinning the publications submitted here originated from my desire not to stop questioning.

Of course, my development was not limited to conceptualisation of good research ideas. I used the opportunities available to develop across all areas as an academic researcher. In Table 6-1, I also show that over the period of the studies, I developed quite considerable skills to position myself towards internationally and publicly renowned publications. I aligned my development with the UK workforce Researcher

Development Framework (RDF) for researcher development. My primary goal for this initiative was to enhance my capacity to develop as a world-class researcher. I therefore took the initiatives to lead and manage the activities to annotate and produce metadata for initial use and later reuse. Some of the activities were more complex, for example, the curation, management and publication of raw data on public databases such as GEO required several standards to be met. Each stage of the material preparation was associated with critical learning that I found extremely helpful.

Similarly, I performed all statistical, mathematical, computational, and bioinformatics analyse or syntheses outlined in the studies underpinning the publications presented here. These processes involved determination and openness to extend my reach

121 and develop high throughput skills in computational biology and statistical analyses.

Consequently, I over the period of investigations improved my knowledge, skills and capabilities in areas beyond biology, genetics and genomics to computer programming, data mining, complex statistics and bioinformatics. These culminated in the design and development of methodology; the creation of novel models, leading and conducting high throughput research and investigation, and in performing both wet and dry-laboratory experiments. Clearly, I received support and guidance and I made a very conscious effort to use every opportunity offered to develop as an all- round researcher in genomics and bioinformatics with core knowledge in human reproduction; and thus, actively contributed to all aspects of the research activities required for recognition as a publisher.

Also, I made conscious efforts during the studies to recognise, create and confidently act on opportunities with the potential to develop my career within both academia and industry. Therefore, I made it a career ambition to actively create and champion opportunities for others within my scope of practice. Hence, I developed a sense of responsiveness towards collaborative opportunities across research disciplines and with non-academic organisations. While I sought to develop substantial ability and confidence to set expectations, I also developed the courage and competence to advise peers and less experienced members of staff to engage in research and scholarly publications. In brief, the journey to develop as a research practitioner has been onerous but worthwhile. I still aspire to develop further to direct not just local policy, but enhance my contributions to shape wider policy and procedures within the

HE sector and professional associations or bodies because human development is intrinsically linked to creation of new knowledge emanating from prudent questioning, and I believe I can still contribute.

122

6.3 Suggestions for future developments The findings from the preceding studies have provided substantial insight into the roles of histamine in the human placenta and the ensuing effect on PE pathophysiology. While the studies have provided closure to some questions, they have also opened up novel areas for further investigations. In this section I outline new areas of investigations that have arisen from above studies.

6.3.1 Early Maternal Blood Pregnancy Histamine and links with Preeclampsia In chapter 1, it was shown that PE is a major cause of perinatal mortality and it complicates up to 8% of all pregnancies in Western countries. In chapter 2, it was also shown that mothers who develop PE present with clinical symptoms akin to experimentally induced elevated histamine and the elevated histamine is harmful to the placenta. While it is widely accepted that the levels and activity of DAO in maternal blood rise a 100 and 1000 fold respectively during normal pregnancy, but diminishes after gestational week 10 in pregnancies that result in PE, the cause of the diminished DAO levels and activity in early pregnancy is unclear. It was further shown in chapter 4 that histamine production is increased in PE placentae, and increasing histamine levels diminishes DAO mRNA expression in placentae. Figure

2-4 further demonstrates that while maternal blood histamine levels in normal pregnancy and late PE pregnancy are well characterised, early maternal blood histamine levels in women who later develop PE remain unelucidated. Therefore, we are currently testing the hypothesis that maternal blood histamine in early pregnancy

(first trimester) are elevated above normal pregnancy levels and the elevated histamine in early pregnancy is associated with diminished maternal serum DAO levels in mothers who later develop PE.

123 We are thus conducting a prospective association study to examine serial assay of maternal blood for histamine and serum for DAO levels of (1) women with no history of PE, (2) women with history of PE and (3) primigravid women with history of histamine intolerance. We anticipate that the findings from this investigation could provide evidence to show for the first time, differences in early pregnancy maternal blood histamine in mothers who later develop PE and those that do not. The findings are also expected to help determine the levels of maternal blood histamine that coincide with diminished maternal serum DAO. Thirdly, with predictive modelling the findings are also expected to suggest maternal histamine and DAO level cut-off points that could be used for early detection of PE.

6.3.2 Histamine Intolerance and Pregnancy Outcome It was shown through the systematic reviews that maternal blood histamine is well controlled in normal pregnancy, and in such instances atopic women who usually have histamine intolerance symptoms fare well during normal pregnancy. In contrast, women whose blood histamine levels are poorly controlled during pregnancy or plausibly elevated from early pregnancy suffer from PE and experience most of the symptoms associated with histamine intolerance. The commentaries summarised the evidence that high blood and urine histamine occur in PE. It is established that placental HDC activity is increased while DAO activity is decreased in PE. It is also established that histamine intolerance can be triggered by certain foods such as beans and pulses, tomatoes, canned fish, cheese, cocoa and chocolate and other histamine releasing foods.

124 It is further established that maternal nutrient deficiency in copper and ascorbic acid increases maternal risk for PE. Copper and ascorbic acid are co-factors for histamine synthesis and metabolism. However, it is not known whether histamine releasing foods or deficiencies in copper and ascorbic acid have effects on the diminished DAO activity in maternal blood. It also uncertain which foods induce histamine elevation in pregnancy so that pregnant women can avoid such foods. We are therefore conducting further studies in this area to investigate the effect of environmental histamine releasers on pregnancy outcome. Specifically, the studies are focused to:

. prospectively survey the dietary type of a cohort of pregnant women and

correlate with the outcome of pregnancy

. assay blood histamine, ascorbic acid and DAO levels and determine their

relation with dietary intake and pregnancy outcome

. determine frequency of exposure to environmental histamine releasers such

as ‘cigarette smoking’ alcohol, and lack of sleep that induce elevated blood

histamine and how these correlate with symptoms of histamine intolerance

during pregnancy.

These studies will have clear impact on maternal health and pregnancy outcomes.

The information thus gained is expected to provide useful dietary and environmental advice to pregnant women who may be at a higher risk of developing trophoblastic diseases such as preeclampsia. It is also expected that the study will yield results to inform a predictive tool for early diagnosis of preeclampsia.

125 6.3.3 Genetic Variations in histamine pathways and Pregnancy Outcomes

It has been recognised that single nucleotide polymorphisms in key genes in histamine pathways including Methylenetetrahydrofolate reductase (MTHFR), DAO

(also known as AOC1), Monoamine Oxidase (MAO), HNMT, and

Phosphatidylethanolamine N-Methyltransferase (PEMT) are strongly linked with defective regulation of blood histamine with and histamine intolerance. Epigenetic regulation of histamine synthesis and metabolic genes has also been associated with defective histamine metabolism and elevated blood histamine. Although the placenta is the root cause of PE, previous studies examining variations in these genes in relation to PE have focused on the mother instead of the placenta. We have therefore adopted an integrated approach to examine the variations in these genes in the placenta in the context of clinical changes in maternal health. Currently we are focusing on characterising the genetic variations of these genes in placenta with high therapeutic potentials to develop and test a multi-sensory diagnostic intervention for early detection and management of PE.

6.3.4 Validation of HSPE genes and curation of histamine regulated pathways in PE placentae A total of 270 significant genes were identified as consistently expressed, specifically in response to elevated histamine treatment in the placenta. These genes were observed to have functional effects that could have implications in the placental pathophysiology of PE. In order to establish the roles of elevated histamine in PE placental pathophysiology and curate the respective pathways, the expression of the

HSPE genes require further validation assessment. It is therefore proposed to undertake further work using RT-qPCR to confirm the expression of the HSPE

126 genes. This validation study could confirm the effects of elevated histamine in the pathophysiology of PE and authenticate subsequent analysis to curate the pathways enriched by the elevated histamine in PE placentae.

6.3.5 Histamine effects on CD3 signal transduction in placental T and NK cells Regulatory loops between histamine and cytokines in human placenta were confirmed in these studies. INF-gamma, IL-1beta and IL-10 transiently induced histamine production. The histamine thus produced was shown to sustain the production of pro-inflammatory and extra-thymus-like cytokines in placenta.

Prolonged exposure of the placental tissues to higher histamine concentration up- regulates TH-1 like cytokine in human placenta explants. There is also emerging evidence that high histamine concentration up-regulates Th-1 response by reversing monocytes or macrophage down regulation of CD3 signal transduction on T and NK cells thereby allowing IL-2 and other Th-1 cytokines to potently sustain the Th-1 cytolytic response (Hellstrand et al. 2000, Johansson et al. 2000). I would like to conduct further studies on this subject to delineate the relationships between histamine receptors and the regulations of CD3 signal transduction in placental T and NK cells.

127 Reference

Aban, M., Cinel, L., Arslan, M., Dilek, U., Kaplanoglu, M., Arpaci, R. & Dilek, S. (2004) Expression of nuclear factor-kappa B and placental apoptosis in pregnancies complicated with intrauterine growth restriction and preeclampsia: an immunohistochemical study. The Tohoku Journal of Experimental Medicine, 204(0040-8727; 0040-8727; 3), p.195-202. Achari, G., Achari, K. & Rao, K.K. (1971) Histaminase and histamine in normal and toxaemic pregnancy. Jpn.J.Pharmacol., 21(0021-5198; 1), p.33-40. Ahlmark, A. (1944) Studies on Histaminolytic Power of the Plasma with Special Reference to Pregnancy. Acta Physiol Scand., 9. Ahmed, A., Dunk, C., Kniss, D. & Wilkes, M. (1997) Role of VEGF receptor-1 (Flt-1) in mediating calcium-dependent nitric oxide release and limiting DNA synthesis in human trophoblast cells. Lab Invest, 76(0023-6837; 0023-6837; 6), p.779-791. Ahmed, A., Li, X.F., Dunk, C., Whittle, M.J., Rushton, D.I. & Rollason, T. (1995) Colocalisation of vascular endothelial growth factor and its Flt-1 receptor in human placenta. Growth Factors (Chur, Switzerland), 12(0897-7194; 0897-7194; 3), p.235-243. Ahmed, A. & Ramma, W. (2015) Unravelling the theories of pre-eclampsia: are the protective pathways the new paradigm? British Journal of Pharmacology, 172(6), p.1574-1586. Akerman, S., Williamson, D.J., Kaube, H. & Goadsby, P.J. (2002) The role of histamine in dural vessel dilation. Brain Res., 956(0006-8993; 1), p.96-102. Andraweera, P.H., Dekker, G.A. & Roberts, C.T. (2012) The vascular endothelial growth factor family in adverse pregnancy outcomes. Human Reproduction Update, 18(4), p.436-457. Arad, G., Nussinovich, R., Na'amad, M. & Kaempfer, R. (1996) Dual control of human interleukin-2 and interferon-gamma gene expression by histamine: activation and suppression. Cellular Immunology, 170(0008-8749; 0008-8749; 1), p.149-155. Arngrimsson, R., Bjornsson, S., Geirsson, R.T., Bjornsson, H., Walker, J.J. & Snaedal, G. (1990) Genetic and familial predisposition to eclampsia and pre-eclampsia in a defined population. Br.J.Obstet.Gynaecol., 97(0306-5456; 0306-5456; 9), p.762-769. Asemissen, A.M., Scheibenbogen, C., Letsch, A., Hellstrand, K., Thoren, F., Gehlsen, K., Schmittel, A., Thiel, E. & Keilholz, U. (2005) Addition of histamine to interleukin 2 treatment augments type 1 T-cell responses in patients with melanoma in vivo: immunologic results from a randomized clinical trial of interleukin 2 with or without histamine (MP 104). Clinical Cancer Research : An Official Journal of the American Association for Cancer Research, 11(1), p.290-297. Baczyk, D., Dunk, C., Huppertz, B., Maxwell, C., Reister, F., Giannoulias, D. & Kingdom, J.C. (2006) Bi-potential behaviour of cytotrophoblasts in first trimester chorionic villi. Placenta, 27(0143- 4004; 0143-4004; 4-5), p.367-374. Bakker, R.A., Casarosa, P., Timmerman, H., Smit, M.J. & Leurs, R. (2004) Constitutively active Gq/11-coupled receptors enable signaling by co-expressed G(i/o)-coupled receptors. Journal of Biological Chemistry, 279(0021-9258; 0021-9258; 7), p.5152-5161. Beaven, M.A. (1982) Factors Regulating Availability of Histamine at Tissue Receptors. In: C.R.Ganellin & M.E.Parsons (eds.) The Pharmacology of Histamine Receptors.: Butterworth Heinemann., p.103-145. Beaven, M.A. (1978) Histamine: its role in physiological and pathological processes. Monogr Allergy, 13(0077-0760), p.1-113. Beaven, M.A., Marshall, J.R., Baylin, S.B. & Sjoerdsma, A. (1975) Changes in plasma histaminase activity during normal early human pregnancy and pregnancy disorders. Am.J.Obstet.Gynecol., 123(0002-9378; 6), p.605-609. Beaven, M.A. & Shaff, R.E. (1975) Study of the relationship of histaminase and diamine oxidase activities in various rat tissues and plasma by sensitive isotopic assay procedures. Biochem.Pharmacol., 24(0006-2952; 9), p.979-984.

128 Belo, L., Santos-Silva, A., Caslake, M., Cooney, J., Pereira-Leite, L., Quintanilha, A. & Rebelo, I. (2003) Neutrophil activation and C-reactive protein concentration in preeclampsia. Hypertension in Pregnancy, 22(2), p.129-141. Best, C.H. (1929) The disappearance of histamine from autolysing lung tissue. The Journal of Physiology, 67(3), p.256-263. Bjuro, T., Lindberg, S. & Westling, H. (1964) Further Observations on the Urinary Excretion of Histamine during and after Normal Pregnancy. Acta Obstetricia Et Gynecologica Scandinavica, 43(0001-6349), p.206-213. Blalock, E. (2003) Experimental Design and Data Analysis. In: Blalock, E. (ed.) A Beginner's Guide to Microarray. London: Kluwer Academic Publishers., p.179-242. Blanco, Y.S., Rozada, H., Remedio, M.R., Hendricks, C.H. & Alvarez, H. (1970) Human tubal motility in vivo. American Journal of Obstetrics and Gynecology, 106(0002-9378; 1), p.79-86. Bode, C.J., Jin, H., Rytting, E., Silverstein, P.S., Young, A.M. & Audus, K.L. (2006) In vitro models for studying trophoblast transcellular transport. Methods in Molecular Medicine, 122, p.225-239. Brand, A., Allen, L., Altman, M., Hlava, M. & Scott, J. (2015) Beyond authorship: attribution, contribution, collaboration, and credit. Learned Publishing, 28, p.151-155. Breitling, R., Armengaud, P., Amtmann, A. & Herzyk, P. (2004) Rank products: a simple, yet powerful, new method to detect differentially regulated genes in replicated microarray experiments. FEBS Lett., 573(0014-5793; 0014-5793; 1-3), p.83-92. Bronson, R.A. & Wallach, E.E. (1977) Lysis of periadnexal adhesions for correction of infertility. Fertil.Steril., 28(0015-0282; 6), p.613-619. Brunton, L. (1996) Agents Affecting Gastrointestinal Water Flux and Motility; Emesis and Antiemetics; Bile acids and Pancreatic Enzymes. In: Hardman JG, Limbird LE, Molinoff PB, Ruddon RW & Gillman AG (eds.) Goodman and Gillman's The Pharmacological Basis of Therapeutics. New York: McGraw-Hill., p.917-937. Buffoni, F. (1966) Histaminase and related amine oxidases. Pharmacol.Rev., 18(0031-6997; 4), p.1163-1199. Burton, G.J. & Caniggia, I. (2001) Hypoxia: implications for implantation to delivery-a workshop report. Placenta, 22 Suppl A(0143-4004), p.S63-S65. Burton, G.J., Charnock-Jones, D.S. & Jauniaux, E. (2006) Working with oxygen and oxidative stress in vitro. Methods in Molecular Medicine, 122(1543-1894; 1543-1894), p.413-425. Burton, G.J. & Jauniaux, E. (2004) Placental oxidative stress: from miscarriage to preeclampsia. Journal of the Society for Gynecologic Investigation, 11(1071-5576; 1071-5576; 6), p.342-352. Buurma, A.J., Turner, R.J., Driessen, J.H., Mooyaart, A.L., Schoones, J.W., Bruijn, J.A., Bloemenkamp, K.W., Dekkers, O.M. & Baelde, H.J. (2013) Genetic variants in pre-eclampsia: a meta-analysis. Human Reproduction Update, 19(1460-2369; 1355-4786; 3), p.289-303. Bytautiene, E., Romero, R., Vedernikov, Y.P., El Zeky, F., Saade, G.R. & Garfield, R.E. (2004a) Induction of premature labor and delivery by allergic reaction and prevention by histamine H1 receptor antagonist. Am.J.Obstet.Gynecol., 191(0002-9378; 4), p.1356-1361. Bytautiene, E., Vedernikov, Y.P., Saade, G.R., Romero, R. & Garfield, R.E. (2004b) Degranulation of uterine mast cell modifies contractility of isolated myometrium from pregnant women. American Journal of Obstetrics and Gynecology, 191(0002-9378; 5), p.1705-1710. Bytautiene, E., Vedernikov, Y.P., Saade, G.R., Romero, R. & Garfield, R.E. (2003) Effect of histamine on phasic and tonic contractions of isolated uterine tissue from pregnant women. Am.J.Obstet.Gynecol., 188(0002-9378; 3), p.774-778. Caldwell, E.J., Carlson, S.E., Palmer, S.M. & Rhodes, P.G. (1988) Histamine and ascorbic acid: a survey of women in labor at term and significantly before term. Int.J.Vitam.Nutr.Res., 58(0300- 9831; 3), p.319-325. Caniggia, I., Winter, J., Lye, S.J. & Post, M. (2000) Oxygen and placental development during the first trimester: implications for the pathophysiology of pre-eclampsia. Placenta, 21 Suppl A(0143- 4004), p.S25-S30.

129 Castellucci, M. & Kaufmann, P. (1982) A three-dimensional study of the normal human placental villous core: II. Stromal architecture. Placenta, 3(0143-4004; 0143-4004; 3), p.269-285. Castellucci, M., Scheper, M., Scheffen, I., Celona, A. & Kaufmann, P. (1990) The development of the human placental villous tree. Anatomy and Embryology, 181(2), p.117-128. Chambers, J.C., Fusi, L., Malik, I.S., Haskard, D.O., De Swiet, M. & Kooner, J.S. (2001) Association of maternal endothelial dysfunction with preeclampsia. Jama, 285(0098-7484; 12), p.1607-1612. Chatterjee, I.B., Gupta, S.D., Majumder, A.K., Nandi, B.K. & Subramanian, N. (1975) Effect of ascorbic acid on histamine metabolism in scorbutic guinea-pigs. J.Physiol, 251(2), p.271-279. Chesley, L.C. & Cooper, D.W. (1986) Genetics of hypertension in pregnancy: possible single gene control of pre-eclampsia and eclampsia in the descendants of eclamptic women. Br.J.Obstet.Gynaecol., 93(0306-5456; 0306-5456; 9), p.898-908. Cincotta, R.B. & Brennecke, S.P. (1998) Family history of pre-eclampsia as a predictor for pre- eclampsia in primigravidas. Int.J.Gynaecol.Obstet., 60(0020-7292; 0020-7292; 1), p.23-27. Clemetson, C.A. (1980) Histamine and ascorbic acid in human blood. J.Nutr., 110(0022-3166; 4), p.662-668. Clemetson, C.A. & Cafaro, V. (1981) Abruptio placentae. Int.J.Gynaecol.Obstet., 19(0020-7292; 6), p.453-460. Cocchiara, R., Albeggiani, G., Di Trapani, G., Azzolina, A., Lampiasi, N., Rizzo, F., Diotallevi, L., Gianaroli, L. & Geraci, D. (1992) Oestradiol enhances in vitro the histamine release induced by embryonic histamine-releasing factor (EHRF) from uterine mast cells. Hum.Reprod., 7(0268- 1161; 8), p.1036-1041. Cocchiara, R., Di Trapani, G., Azzolina, A., Albeggiani, G., Ciriminna, R., Cefalu, E., Cittadini, E. & Geraci, D. (1987) Isolation of a histamine releasing factor from human embryo culture medium after in-vitro fertilization. Hum.Reprod., 2(0268-1161; 4), p.341-344. Cocchiara, R., Lampiasi, N., Albeggiani, G., Azzolina, A., Bongiovanni, A., Gianaroli, L., Di Blasi, F. & Geraci, D. (1996) A factor secreted by human embryo stimulates cytokine release by uterine mast cell. Mol.Hum.Reprod., 2(1360-9947; 10), p.781-791. Connolly, M.R., Bitman, J., CECIL, H.C. & Wrenn, T.R. (1962) Water, electrolyte, glycogen, and histamine content of rat uterus during pregnancy. Am.J.Physiol, 203(0002-9513), p.717-719. Cooper, J.A. & Schayer, R.W. (1956) Metabolism of C14 histamine in man. J.Appl.Physiol, 9(0021- 8987; 3), p.481-483. Cricco, G., Martin, G., Medina, V., Nunez, M., Mohamad, N., Croci, M., Crescenti, E., Bergoc, R. & Rivera, E. (2006) Histamine inhibits cell proliferation and modulates the expression of Bcl-2 family proteins via the H2 receptor in human pancreatic cancer cells. Anticancer Research, 26(0250-7005; 0250-7005; 6), p.4443-4450. Dale, H.H. & Laidlaw, P.P. (1910) The physiological action of b-imidazolylethylamine. The Journal of Physiology, 41, p.318-344. Damsky, C., Sutherland, A. & Fisher, S. (1993) Extracellular matrix 5: adhesive interactions in early mammalian embryogenesis, implantation, and placentation. Faseb J., 7(0892-6638; 14), p.1320- 1329. Damsky, C., Schick, S.F., Klimanskaya, I., Stephens, L., Zhou, Y. & Fisher, S. (1997) Adhesive interactions in peri-implantation morphogenesis and placentation. Reprod.Toxicol., 11(0890- 6238; 0890-6238; 2-3), p.367-375. Daoud, G., Barrak, J. & Abou-Kheir, W. (2016) Assessment Of Different Trophoblast Cell Lines As In Vitro Models For Placental Development. The FASEB Journal, 30(1 Supplement), p.1247.18- 1247.18. Darmochwal-Kolarz, D., Leszczynska-Gorzelak, B., Rolinski, J. & Oleszczuk, J. (1999) T helper 1- and T helper 2-type cytokine imbalance in pregnant women with pre-eclampsia. European Journal of Obstetrics, Gynecology, and Reproductive Biology, 86(0301-2115; 0301-2115; 2), p.165-170. Davio, C., Mladovan, A., Lemos, B., Monczor, F., Shayo, C., Rivera, E. & Baldi, A. (2002) H1 and H2 histamine receptors mediate the production of inositol phosphates but not cAMP in human breast

130 epithelial cells. Inflammation Research : Official Journal of the European Histamine Research Society ... [Et Al.], 51(1023-3830; 1), p.1-7. Dey, S.K. & Hubbard, C.J. (1981) Role of histamine and cyclic nucleotides in implantation in the rabbit. Cell and Tissue Research, 220(0302-766; 0302-766; 3), p.549-554. Dey, S.K., Lim, H., Das, S.K., Reese, J., Paria, B.C., Daikoku, T. & Wang, H. (2004) Molecular cues to implantation. Endocr.Rev., 25(0163-769; 3), p.341-373. Dobbin, K. & Simon, R. (2005) Sample size determination in microarray experiments for class comparison and prognostic classification. Biostatistics., 6(1465-4644; 1465-4644; 1), p.27-38. Donahue, J.G., Lupton, J.B. & Golichowski, A.M. (1995) Cutaneous mastocytosis complicating pregnancy. Obstet.Gynecol., 85(0029-7844; 5), p.813-815. Dong, M., He, J., Wang, Z., Xie, X. & Wang, H. (2005) Placental imbalance of Th1- and Th2-type cytokines in preeclampsia. Acta Obstetricia Et Gynecologica Scandinavica, 84(8), p.788-793. Dubois, A.M., Santais, M.C., Foussard, C., Dubois, F., Ruff, F., Taurelle, R. & Parrot, J.L. (1977) Blood histamine and plasma histaminase level during human pregnancy [proceedings]. Agents Actions, 7(0065-4299; 1), p.112. Dudley, A.M., Aach, J., Steffen, M.A. & Church, G.M. (2002) Measuring absolute expression with microarrays with a calibrated reference sample and an extended signal intensity range. Proc.Natl.Acad.Sci.U.S.A, 99(0027-8424; 0027-8424; 11), p.7554-7559. Duley, L. (2009) The global impact of pre-eclampsia and eclampsia. Seminars in Perinatology, 33(3), p.130-137. Ebeigbe, A.B. & Talabi, O.O. (2014) Vascular Effects of Histamine. Nigerian Journal of Physiological Sciences : Official Publication of the Physiological Society of Nigeria, 29(1), p.7-10. Efron, B. & Tibshirani, R. (2007) On testing the significance of sets of genes. Ann.Appl.Stat., 1(1), p.107-129. el-Gendi, M.A., Nassar, S.H. & el-Bassiouni, E.A. (1988) Histamine and histamine receptor antagonists in splanchnic compartments in schistosomal portal hypertension. Clinical Pharmacology and Therapeutics, 44(4), p.396-399. Esplin, M.S., Fausett, M.B., Fraser, A., Kerber, R., Mineau, G., Carrillo, J. & Varner, M.W. (2001) Paternal and maternal components of the predisposition to preeclampsia. N.Engl.J.Med., 344(0028-4793; 0028-4793; 12), p.867-872. Fabregat, A., Sidiropoulos, K., Garapati, P., Gillespie, M., Hausmann, K., Haw, R., Jassal, B., Jupe, S., Korninger, F., McKay, S., Matthews, L., May, B., Milacic, M., Rothfels, K., Shamovsky, V., Webber, M., Weiser, J., Williams, M., Wu, G., Stein, L., Hermjakob, H. & D'Eustachio, P. (2016) The Reactome pathway Knowledgebase. Nucleic Acids Research, 44(D1), p.D481-7. Fajardy, I., Moitrot, E., Vambergue, A., Vandersippe-Millot, M., Deruelle, P. & Rousseaux, J. (2009) Time course analysis of RNA stability in human placenta. BMC Molecular Biology [Computer File], 10(1471-2199; 1471-2199), p.21. Falus, A. & Meretey, K. (1992) Histamine: an early messenger in inflammatory and immune reactions. Immunology Today, 13(0167-5699; 5), p.154-156. Faye, A., Pornprasert, S., Dolcini, G., Ave, P., Taieb, J., Taupin, J.L., Derrien, M., Huerre, M., Barre- Sinoussi, F., Chaouat, G. & Menu, E. (2005) Evaluation of the placental environment with a new in vitro model of histocultures of early and term placentae: determination of cytokine and chemokine expression profiles. Placenta, 26(0143-4004; 0143-4004; 2-3), p.262-267. Fisher, S.J. & Damsky, C.H. (1993) Human cytotrophoblast invasion. Seminars in Cell Biology, 4(1043-4682; 3), p.183-188. Fukuda, M., Tanaka, S., Suzuki, S., Kusama, K., Kaneko, T. & Sakashita, H. (2007) Cimetidine induces apoptosis of human salivary gland tumor cells. Oncology Reports, 17(1021-335; 1021- 335; 3), p.673-678. Gammill, H.S. & Roberts, J.M. (2007) Emerging concepts in preeclampsia investigation. Frontiers in Bioscience : A Journal and Virtual Library, 12, p.2403-2411.

131 Genbacev, O., Joslin, R., Damsky, C.H., Polliotti, B.M. & Fisher, S.J. (1996) Hypoxia alters early gestation human cytotrophoblast differentiation/invasion in vitro and models the placental defects that occur in preeclampsia. J.Clin.Invest, 97(0021-9738; 2), p.540-550. Genbacev, O. & Miller, R.K. (2000) Post-implantation differentiation and proliferation of cytotrophoblast cells: in vitro models--a review. Placenta, 21 Suppl A(0143-4004), p.S45-S49. Ghosh, A.K., Hirasawa, N. & Ohuchi, K. (2001) Enhancement by histamine of vascular endothelial growth factor production in granulation tissue via H(2) receptors. British Journal of Pharmacology, 134(0007-1188; 0007-1188; 7), p.1419-1428. Gierman, L.M., Stodle, G.S., Tangeras, L.H., Austdal, M., Olsen, G.D., Follestad, T., Skei, B., Rian, K., Gundersen, A.S., Austgulen, R. & Iversen, A.C. (2015) Toll-like receptor profiling of seven trophoblast cell lines warrants caution for translation to primary trophoblasts. Placenta, 36(11), p.1246-1253. Greer, I.A., Dawes, J., Johnston, T.A. & Calder, A.A. (1991) Neutrophil activation is confined to the maternal circulation in pregnancy-induced hypertension. Obstetrics and Gynecology, 78(1), p.28- 32. Gunther, R.E. & Glick, D. (1967) Determination of histaminase activity in histologic samples and its quantitative distribution in intact human placenta and uterus. J.Histochem.Cytochem., 15(0022- 1554; 8), p.431-435. Hamano, N., Terada, N., Maesako, K., Ikeda, T., Fukuda, S., Wakita, J., Yamashita, T. & Konno, A. (1998) Expression of histamine receptors in nasal epithelial cells and endothelial cells--the effects of sex hormones. Int.Arch.Allergy Immunol., 115(1018-2438; 3), p.220-227. Harris, L.K., Smith, S.D., Keogh, R.J., Jones, R.L., Baker, P.N., Knofler, M., Cartwright, J.E., Whitley, G.S. & Aplin, J.D. (2010) Trophoblast- and vascular smooth muscle cell-derived MMP-12 mediates elastolysis during uterine spiral artery remodeling. American Journal of Pathology, 177(1525-2191; 0002-9440; 4), p.2103-2115. Hellstrand, K., Brune, M., Naredi, P., Mellqvist, U.H., Hansson, M., Gehlsen, K.R. & Hermodsson, S. (2000) Histamine: a novel approach to cancer immunotherapy. Cancer Investigation, 18(0735- 7907; 4), p.347-355. Hill, S.J. (1990) Distribution, properties, and functional characteristics of three classes of histamine receptor. Pharmacol.Rev., 42(0031-6997; 1), p.45-83. Hine, R.J., Orsini, M.W. & Hegstrand, L.R. (1985) Changes in tissue histamine during the estrous cycle, pregnancy and pseudopregnancy in the golden hamster. Proc.Soc.Exp.Biol.Med., 179(0037-9727; 3), p.271-278. Hong, F. & Breitling, R. (2008) A comparison of meta-analysis methods for detecting differentially expressed genes in microarray experiments. Bioinformatics, 24(1367-4811; 1367-4803; 3), p.374-382. Horakova, Z., Keiser, H.R. & Beaven, M.A. (1977) Blood and urine histamine levels in normal and pathological states as measured by a radiochemical assay. Clinica Chimica Acta; International Journal of Clinical Chemistry, 79(0009-8981; 2), p.447-456. Huppertz, B., Kingdom, J., Caniggia, I., Desoye, G., Black, S., Korr, H. & Kaufmann, P. (2003) Hypoxia favours necrotic versus apoptotic shedding of placental syncytiotrophoblast into the maternal circulation. Placenta, 24(0143-4004; 2-3), p.181-190. Ilekis, J.V., Tsilou, E., Fisher, S., Abrahams, V.M., Soares, M.J., Cross, J.C., Zamudio, S., Illsley, N.P., Myatt, L., Colvis, C., Costantine, M.M., Haas, D.M., Sadovsky, Y., Weiner, C., Rytting, E. & Bidwell, G. (2016) Placental origins of adverse pregnancy outcomes: potential molecular targets: an Executive Workshop Summary of the Eunice Kennedy Shriver National Institute of Child Health and Human Development. American Journal of Obstetrics and Gynecology, . Ind, P.W., Barnes, P.J., Brown, M.J., Causon, R. & Dollery, C.T. (1983) Measurement of plasma histamine in asthma. Clin.Allergy, 13(0009-9090; 1), p.61-67. Ind, P.W., Brown, M.J., Lhoste, F.J., Macquin, I. & Dollery, C.T. (1982) Concentration effect relationships of infused histamine in normal volunteers. Agents Actions, 12(0065-4299; 1-2), p.12-16.

132 Irving, J.A., Lysiak, J.J., Graham, C.H., Hearn, S., Han, V.K. & Lala, P.K. (1995) Characteristics of trophoblast cells migrating from first trimester chorionic villus explants and propagated in culture. Placenta, 16(0143-4004; 0143-4004; 5), p.413-433. James, J.L., Stone, P.R. & Chamley, L.W. (2006) The effects of oxygen concentration and gestational age on extravillous trophoblast outgrowth in a human first trimester villous explant model. Hum.Reprod., 21(0268-1161; 0268-1161; 10), p.2699-2705. James, J.L., Stone, P.R. & Chamley, L.W. (2005) Cytotrophoblast differentiation in the first trimester of pregnancy: evidence for separate progenitors of extravillous trophoblasts and syncytiotrophoblast. Reproduction (Cambridge, England), 130(1470-1626; 1470-1626; 1), p.95- 103. Jankovic, S.M., Varjacic, M. & Jankovic, S.V. (1998) Different roles of histamine receptor subtypes in ampullar & isthmic segments of human fallopian tube. Indian J.Med.Res., 107(0971-5916), p.224-230. Johansson, M., Henriksson, R., Bergenheim, A.T. & Koskinen, L.O. (2000) Interleukin-2 and histamine in combination inhibit tumour growth and angiogenesis in malignant glioma. British Journal of Cancer, 83(6), p.826-832. Johnson, D.C. & Dey, S.K. (1980) Role of histamine in implantation: dexamethasone inhibits estradiol-induced implantation in the rat. Biol.Reprod., 22(0006-3363; 5), p.1136-1141. Juran, J.M. (1975) The non-Pareto principle; mea culpa. Juran’s Quality, , p.1021. Jutel, M., Watanabe, T., Klunker, S., Akdis, M., Thomet, O.A., Malolepszy, J., Zak-Nejmark, T., Koga, R., Kobayashi, T., Blaser, K. & Akdis, C.A. (2001) Histamine regulates T-cell and antibody responses by differential expression of H1 and H2 receptors. Nature, 413(0028-0836; 0028- 0836; 6854), p.420-425. Kahlson, G., Nilsson, K., Rosengren, E. & Zederfeldt, B. (1960a) Wound healing as on rate of histamine formation. Lancet, 2, p.230-234. Kahlson, G. & Rosengren, E. (1971) Biogenesis and physiology of histamine. Monogr Physiol Soc., 21(0079-2020), p.1-318. Kahlson, G., Rosengren, E. & WESTLING, H. (1958a) Increased formation of histamine in the pregnant rat. J.Physiol, 143(0022-3751; 1), p.91-103. Kahlson, G., Rosengren, E., WESTLING, H. & WHITE, T. (1958b) The site of increased formation of histamine in the pregnant rat. J.Physiol, 144(0022-3751; 2), p.337-348. Kahlson, G., Rosengren, E. & WHITE, T. (1960b) The formation of histamine in the rat foetus. J.Physiol, 151, p.131-138. Kaliner, M., Shelhamer, J.H. & Ottesen, E.A. (1982) Effects of infused histamine: correlation of plasma histamine levels and symptoms. J.Allergy Clin.Immunol., 69(0091-6749; 3), p.283-289. Kambayashi, T. & Koretzky, G.A. (2007) Proximal signaling events in Fc epsilon RI-mediated mast cell activation. The Journal of Allergy and Clinical Immunology, 119(3), p.544-52; quiz 553-4. Kapeller-Adler, R. (1965) Histamine Catabolism in vitro and in vivo. Federation Proceedings, 24(0014- 9446), p.757-765. Kapeller-Adler, R. (1952a) [Metabolism of histidine and histamine in normal and toxemic pregnancy]. Brux.Med., 32(0068-3027; 31), p.1601-1617. Kapeller-Adler, R. (1952b) Histamine metabolism in the human placenta and in the umbilical cord blood. Biochem.J., 51(0264-6021; 5), p.610-613. Kapeller-Adler, R. (1949) Histamine metabolism in pregnancy. Lancet, 2(0140-6736; 17), p.745-747. Kapeller-Adler, R. (1944) Investigation on the activity of histaminase in normal and toxaemic pregnancy. The Biochemical Journal, 38, p.270-274. Kapeller-Adler, R. (1941) Significance of Isolation of Histamine from the urine in Toxaemia of Pregnancy. Journal of Obstetrics and Gynaecology (Tokyo, Japan), 48, p.155-160. Kauma, S.W., Walsh, S.W., Nestler, J.E. & Turner, T.T. (1992) Interleukin-1 is induced in the human placenta by endotoxin and isolation procedures for trophoblasts. J.Clin.Endocrinol.Metab, 75(0021-972; 3), p.951-955.

133 Kerr, M.K. & Churchill, G.A. (2001) Statistical design and the analysis of gene expression microarray data. Genetical Research, 77(0016-6723; 0016-6723; 2), p.123-128. Khan, K.S., Wojdyla, D., Say, L., Gulmezoglu, A.M. & Van Look, P.F. (2006) WHO analysis of causes of maternal death: a systematic review. Lancet (London, England), 367(9516), p.1066-1074. Khong, T.Y., Tee, J.H. & Kelly, A.J. (1997) Absence of innervation of the uteroplacental arteries in normal and abnormal human pregnancies. Gynecologic and Obstetric Investigation, 43(2), p.89- 93. King, A.E., Critchley, H.O. & Kelly, R.W. (2003) Innate immune defences in the human endometrium. Reproductive Biology and Endocrinology : RB&E, 1, p.116. Kirkel', A.Z., Pekkel', V.A., Akhmetalieva, R.Z., Chernukha, E.A. & Gorkin, V.Z. (1983) [Decrease of placental amine oxidase activities in premature birth]. Vopr.Med.Khim., 29(0042-8809; 4), p.83- 87. Kramer, M.S., Coates, A.L., Michoud, M.C., Dagenais, S., Moshonas, D., Davis, G.M., Hamilton, E.F., Nuwayhid, B., Joshi, A.K., Papageorgiou, A. & . (1995) Maternal asthma and idiopathic preterm labor. Am.J.Epidemiol., 142(0002-9262; 10), p.1078-1088. Kubo, Y. & Nakano, K. (1999) Regulation of histamine synthesis in mouse CD4+ and CD8+ T lymphocytes. Inflammation Research : Official Journal of the European Histamine Research Society ... [Et Al.], 48(1023-3830; 3), p.149-153. Lassen, L.H., Christiansen, I., Iversen, H.K., Jansen-Olesen, I. & Olesen, J. (2003) The effect of nitric oxide synthase inhibition on histamine induced headache and arterial dilatation in migraineurs. Cephalalgia : An International Journal of Headache, 23(0333-1024; 9), p.877-886. Leach, L., Eaton, B.M., Westcott, E.D. & Firth, J.A. (1995) Effect of histamine on endothelial permeability and structure and adhesion molecules of the paracellular junctions of perfused human placental microvessels. Microvasc.Res., 50(3), p.323-337. Leach, L. & Firth, J.A. (1997) Structure and permeability of human placental microvasculature. Microsc.Res.Tech., 38(1059-910; 1-2), p.137-144. Lee, M.L., Kuo, F.C., Whitmore, G.A. & Sklar, J. (2000) Importance of replication in microarray gene expression studies: statistical methods and evidence from repetitive cDNA hybridizations. Proceedings of the National Academy of Sciences of the United States of America, 97(0027- 8424; 0027-8424; 18), p.9834-9839. Legge, M. & Duff, G.B. (1981) Plasma diamine oxidase levels in pregnancy complicated by threatened abortion. Journal of Clinical Pathology, 34(0021-9746; 2), p.187-188. Lewis, M.P., Clements, M., Takeda, S., Kirby, P.L., Seki, H., Lonsdale, L.B., Sullivan, M.H., Elder, M.G. & White, J.O. (1996) Partial characterization of an immortalized human trophoblast cell- line, TCL-1, which possesses a CSF-1 autocrine loop. Placenta, 17(0143-4004; 2-3), p.137-146. Lim, K.J., Odukoya, O.A., Ajjan, R.A., Li, T.C., Weetman, A.P. & Cooke, I.D. (2000) The role of T- helper cytokines in human reproduction. Fertil.Steril., 73(0015-0282; 0015-0282; 1), p.136-142. Lindberg, S. (1963a) 14-C-histamine elimination from the blood of pregnant and non-pregnant women with special reference to the uterus. Acta Obstetricia Et Gynecologica Scandinavica. Supplement, 42(Suppl 1)(0300-8835), p.3-25. Lindberg, S. (1963b) 14-C-histamine inactivation in vitro by human myometrial and placental tissues. Acta Obstetricia Et Gynecologica Scandinavica. Supplement, 42(Suppl 1)(0300-8835), p.26-34. Lindberg, S., LINDELL, S.E. & WESTLING, H. (1963a) Formation and inactivation of histamine by human foetal tissues in vitro. Acta Obstetricia Et Gynecologica Scandinavica. Supplement, 42(Suppl 1)(0300-8835), p.49-58. Lindberg, S., LINDELL, S.E. & WESTLING, H. (1963b) The metabolism of 14-C-labelled histamine injected into the umbilical artery. Acta Obstetricia Et Gynecologica Scandinavica. Supplement, 42(Suppl 1)(0300-8835), p.35-47. Lindberg, S. & Tornqvist, A. (1966) The inhibitory effect of aminoguanidine on histamine catabolism in human pregnancy. Acta Obstetricia Et Gynecologica Scandinavica, 45(0001-6349; 2), p.131- 139.

134 Linzke, N., Schumacher, A., Woidacki, K., Croy, B.A. & Zenclussen, A.C. (2014) Carbon monoxide promotes proliferation of uterine natural killer cells and remodeling of spiral arteries in pregnant hypertensive heme oxygenase-1 mutant mice. Hypertension, 63(3), p.580-588. Liu, Z., Kilburn, B.A., Leach, R.E., Romero, R., Paria, B.C. & Armant, D.R. (2004) Histamine enhances cytotrophoblast invasion by inducing intracellular calcium transients through the histamine type-1 receptor. Molecular Reproduction and Development, 68(1040-452; 1040-452; 3), p.345-353. Lonsdale, L.B., Elder, M.G. & Sullivan, M.H. (1996) A comparison of cytokine and hormone production by decidual cells and tissue explants. J.Endocrinol., 151(0022-0795; 2), p.309-313. Lorenz, M.O. (1905) Methods of Measuring the Concentration of Wealth. American Statistical Association Publication, 9(70), p.200-219. Lorenz, W., Doenicke, A., Meyer, R., Reimann, H.J., Kusche, J., Barth, H., Geesing, H., Hutzel, M. & Weissenbacher, B. (1972) An improved method for the determination of histamine release in man: its application in studies with propanidid and thiopentone. Eur.J.Pharmacol., 19(0014-2999; 2), p.180-190. Lyall, F. (2006) Mechanisms regulating cytotrophoblast invasion in normal pregnancy and pre- eclampsia. The Australian & New Zealand Journal of Obstetrics & Gynaecology, 46(0004-8666; 0004-8666; 4), p.266-273. Lyall, F. (2003) Development of the utero-placental circulation: the role of carbon monoxide and nitric oxide in trophoblast invasion and spiral artery transformation. Microscopy Research and Technique, 60(1059-910; 1059-910; 4), p.402-411. Lyall, F. (2002) The human placental bed revisited. Placenta, 23(0143-4004; 0143-4004; 8-9), p.555- 562. Lynch, A.M., Murphy, J.R., Byers, T., Gibbs, R.S., Neville, M.C., Giclas, P.C., Salmon, J.E. & Holers, V.M. (2008) Alternative complement pathway activation fragment Bb in early pregnancy as a predictor of preeclampsia. American Journal of Obstetrics and Gynecology, 198(4), p.385.e1- 385.e9. Ma, R.Z., Gao, J., Meeker, N.D., Fillmore, P.D., Tung, K.S., Watanabe, T., Zachary, J.F., Offner, H., Blankenhorn, E.P. & Teuscher, C. (2002) Identification of Bphs, an autoimmune disease locus, as histamine receptor H1. Science, 297(1095-9203; 0036-8075; 5581), p.620-623. Marcou, I., Athanasiu-Vergh, E., Chiriceanu, D., Cosma, G., Gingold, N. & Parhon, C.C. (1938) Sur le rôle physiologique de l'histamine. Presse Med., 46, p.371. Martin, C.B.,Jr (1965) Uterine Blood Flow and Placental Circulation. Anesthesiology, 26(0003-3022; 0003-3022), p.447-459. Martinez-Varea, A., Pellicer, B., Perales-Marin, A. & Pellicer, A. (2014) Relationship between maternal immunological response during pregnancy and onset of preeclampsia. Journal of Immunology Research, 2014, p.210241. Marzioni, D., Tamagnone, L., Capparuccia, L., Marchini, C., Amici, A., Todros, T., Bischof, P., Neidhart, S., Grenningloh, G. & Castellucci, M. (2004) Restricted innervation of uterus and placenta during pregnancy: evidence for a role of the repelling signal Semaphorin 3A. Developmental Dynamics : An Official Publication of the American Association of Anatomists, 231(4), p.839-848. Matsubara, M., Ohmori, K. & Hasegawa, K. (2006) Histamine H1 receptor-stimulated interleukin 8 and granulocyte macrophage colony-stimulating factor production by bronchial epithelial cells requires extracellular signal-regulated kinase signaling via protein kinase C. Int.Arch.Allergy Immunol., 139(1018-2438; 1018-2438; 4), p.279-293. Matsuo, K., Kooshesh, S., Dinc, M., Sun, C.C., Kimura, T. & Baschat, A.A. (2007) Late postpartum eclampsia: report of two cases managed by uterine curettage and review of the literature. Am.J.Perinatol., 24(0735-1631; 0735-1631; 4), p.257-266. Matsuyama, K., Ichikawa, T., Nitta, Y., Ikoma, Y., Ishimura, K., Horio, S. & Fukui, H. (2006) Localized expression of histamine H1 receptors in syncytiotrophoblast cells of human placenta. Journal of Pharmacological Sciences, 102(3), p.331-337.

135 Matsuyama, K., Kawakami, N., Ichikawa, T., Nitta, Y., Ishimura, K., Horio, S. & Fukui, H. (2004) Expression of histamine H(1) receptor in placenta. Inflammation Research : Official Journal of the European Histamine Research Society ...[Et Al.], 53 Suppl 1, p.S85-6. McMillan, J.C., Heskel, N.S. & Hanifin, J.M. (1985) Cyclic AMP-phosphodiesterase activity and histamine release in cord blood leukocyte preparations. Acta Derm.Venereol.Suppl (Stockh), 114(0365-8341), p.24-32. Meekins, J.W., Pijnenborg, R., Hanssens, M., McFadyen, I.R. & van, A.A. (1994) A study of placental bed spiral arteries and trophoblast invasion in normal and severe pre-eclamptic pregnancies. British Journal of Obstetrics and Gynaecology, 101(0306-5456; 0306-5456; 8), p.669-674. Menezo, Y.J., Nicollet, B., Dumont, M., Hazout, A. & Janny, L. (1993) Factors affecting human blastocyst formation in vitro and freezing at the blastocyst stage. Acta Europaea Fertilitatis, 24(0587-2421; 0587-2421; 5), p.207-213. Miller, R.K., Genbacev, O., Turner, M.A., Aplin, J.D., Caniggia, I. & Huppertz, B. (2005) Human placental explants in culture: approaches and assessments. Placenta, 26(0143-4004; 6), p.439- 448. Minagawa, M., Narita, J., Tada, T., Maruyama, S., Shimizu, T., Bannai, M., Oya, H., Hatakeyama, K. & Abo, T. (1999) Mechanisms underlying immunologic states during pregnancy: possible association of the sympathetic nervous system. Cell Immunol., 196(1), p.1-13. Mohan, A.R., Sooranna, S.R., Lindstrom, T.M., Johnson, M.R. & Bennett, P.R. (2007) The effect of mechanical stretch on cyclooxygenase type 2 expression and activator protein-1 and nuclear factor-kappaB activity in human amnion cells. Endocrinology, 148(0013-7227; 0013-7227; 4), p.1850-1857. Mol, B.W., Roberts, C.T., Thangaratinam, S., Magee, L.A., de Groot, C.J. & Hofmeyr, G.J. (2016) Pre- eclampsia. Lancet (London, England), 387(10022), p.999-1011. Mondovi, B., Scioscia, S.A., Rotilio, G. & Costa, M.T. (1965) Studies on the Inhibition of Histaminase Activity by Histamine. Enzymologia, 28, p.228-234. Morgan, D.M. & Hytten, F.E. (1984) Mid-trimester pregnancy--a time of tranquility or activity? British Journal of Obstetrics and Gynaecology, 91(0306-5456; 6), p.532-537. Murakami, H., Sun-Wada, G.H., Matsumoto, M., Nishi, T., Wada, Y. & Futai, M. (1999) Human histamine H2 receptor gene: multiple transcription initiation and tissue-specific expression. FEBS Letters, 451(3), p.327-331. Myatt, L. (2002) Role of placenta in preeclampsia. Endocrine, 19(1355-008; 1355-008; 1), p.103-111. Naicker, T., Khedun, S.M., Moodley, J. & Pijnenborg, R. (2003) Quantitative analysis of trophoblast invasion in preeclampsia. Acta Obstetricia Et Gynecologica Scandinavica, 82(0001-6349; 0001- 6349; 8), p.722-729. Nandi, B.K., Subramanian, N., Majumder, A.K. & Chatterjee, I.B. (1974) Effect of ascorbic acid on detoxification of histamine under stress conditions. Biochem.Pharmacol., 23(0006-2952; 3), p.643-647. Nelson, D.M., Meister, R.K., Ortman-Nabi, J., Sparks, S. & Stevens, V.C. (1986) Differentiation and secretory activities of cultured human placental cytotrophoblast. Placenta, 7(0143-4004; 0143- 4004; 1), p.1-16. Neugebauer, E. & Lorenz, W. (1982) A modified Schayer procedure for the estimation of histidine decarboxylase activity: its application on tissue extracts from gastric mucosa of various mammals. Agents Actions, 12(0065-4299; 1-2), p.32-40. NICE (2017) Pre-eclampsia. In: Hypertension in pregnancy pathway.: National Institute for Health and Care Excellence. NICE (2011) Hypertension in pregnancy: diagnosis and management. NICE clinical guideline 107. :National Institute for Health and Clinical Excellence. NICE (2008) Antenatal care for uncomplicated pregnancies. Last updated: March 2016. London :National Institute of Clinical Exellence.

136 Nilsson, E., Salonen, R.H., Cnattingius, S. & Lichtenstein, P. (2004) The importance of genetic and environmental effects for pre-eclampsia and gestational hypertension: a family study. BJOG : An International Journal of Obstetrics and Gynaecology, 111(1470-0328; 1470-0328; 3), p.200-206. Nilsson, G., Lindell, S.E., Schayer, R.W. & Westling, H. (1959) Metabolism of 14C-labelled histamine in pregnant and non-pregnant women. Clin.Sci., 18(0009-9287), p.313-319. Norn, S. (1968) Antigenic histamine release from fractionated and unfractionated peritoneal cells from sensitized rats. Acta Pharmacologica Et Toxicologica, 26(4), p.373-383. Norwitz, E.R., Schust, D.J. & Fisher, S.J. (2001) Implantation and the survival of early pregnancy. The New England Journal of Medicine, 345(0028-4793; 0028-4793; 19), p.1400-1408. Orendi, K., Kivity, V., Sammar, M., Grimpel, Y., Gonen, R., Meiri, H., Lubzens, E. & Huppertz, B. (2011) Placental and trophoblastic in vitro models to study preventive and therapeutic agents for preeclampsia. Placenta, 32 Suppl(1532-3102; 0143-4004), p.S49-S54. Packard, K.A. & Khan, M.M. (2003) Effects of histamine on Th1/Th2 cytokine balance. International Immunopharmacology, 3(1567-5769; 7), p.909-920. Palmer, M.E., Watson, A.L. & Burton, G.J. (1997) Morphological analysis of degeneration and regeneration of syncytiotrophoblast in first trimester placental villi during organ culture. Hum.Reprod., 12(0268-1161; 0268-1161; 2), p.379-382. Panina-Bordignon, P., Mazzeo, D., Lucia, P.D., D'Ambrosio, D., Lang, R., Fabbri, L., Self, C. & Sinigaglia, F. (1997) Beta2-agonists prevent Th1 development by selective inhibition of interleukin 12. The Journal of Clinical Investigation, 100(6), p.1513-1519. Pap, E., Falus, A., Mihalyi, D., Borck, H., Diel, F. & Pallinger, E. (2007) Histamine regulates placental cytokine expression--in vivo study on HDC knockout mice. Placenta, 28(0143-4004; 0143-4004; 2-3), p.239-244. Peng, X. & Stromberg, A.J. (2003) Microarray Experiment Design and Statistical Analysis. In: A Beginner's Guide to Microarray. London: Kluwer Academic Publishers., p.243-275. Pfeiffer, C.C., Iliev, V., Goldstein, L., Jenney, E.H. & Schultz, R. (1970) Blood histamine, polyamines and the schizophrenias. Computer correlations of the low and high blood histamine types. Research Communications in Chemical Pathology and Pharmacology, 1(2), p.247-265. Pijnenborg, R. (1988) Establishment of uteroplacental circulation. Reproduction, Nutrition, Development, 28(0181-1916; 0181-1916; 6), p.1581-1586. Pijnenborg, R., Ball, E., Bulmer, J.N., Hanssens, M., Robson, S.C. & Vercruysse, L. (2006a) In vivo analysis of trophoblast cell invasion in the human. Methods in Molecular Medicine, 122(1543- 1894; 1543-1894), p.11-44. Pijnenborg, R., Vercruysse, L. & Hanssens, M. (2006b) The uterine spiral arteries in human pregnancy: facts and controversies. Placenta, 27(0143-4004; 0143-4004; 9-10), p.939-958. Pollock, I., Murdoch, R.D. & Lessof, M.H. (1991) Plasma histamine and clinical tolerance to infused histamine in normal, atopic and urticarial subjects. Agents Actions, 32(0065-4299; 3-4), p.359- 365. Prieto, J.A., Panyutich, A.V. & Heine, R.P. (1997) Neutrophil activation in preeclampsia. Are defensins and lactoferrin elevated in preeclamptic patients? The Journal of Reproductive Medicine, 42(1), p.29-32. Pringle, K.G., Kind, K.L., Sferruzzi-Perri, A.N., Thompson, J.G. & Roberts, C.T. (2010) Beyond oxygen: complex regulation and activity of hypoxia inducible factors in pregnancy. Human Reproduction Update, 16(1460-2369; 1355-4786; 4), p.415-431. Purcell, W.M. (1992) Human placental mast cells: a role in pre-eclampsia? Med.Hypotheses, 39(0306-9877; 3), p.281-283. Purcell, W.M. & Hanahoe, T.H. (1991) A novel source of mast cells: the human placenta. Agents Actions, 33(0065-4299; 1-2), p.8-12. Qin, S.Q., Kusuma, G.D., Al-Sowayan, B., Pace, R.A., Isenmann, S., Pertile, M.D., Gronthos, S., Abumaree, M.H., Brennecke, S.P. & Kalionis, B. (2016) Establishment and characterization of foetal and maternal mesenchymal stem/stromal cell lines from the human term placenta. Placenta, 39, p.134-146.

137 Raghupathy, R. (2001) Pregnancy: success and failure within the Th1/Th2/Th3 paradigm. Seminars in Immunology, 13(4), p.219-227. Ramma, W., Buhimschi, I.A., Zhao, G., Dulay, A.T., Nayeri, U.A., Buhimschi, C.S. & Ahmed, A. (2012) The elevation in circulating anti-angiogenic factors is independent of markers of neutrophil activation in preeclampsia. Angiogenesis, 15(3), p.333-340. Redman, C. (2014) Pre-eclampsia: A complex and variable disease. Pregnancy Hypertension, 4(3), p.241-242. Redman, C.W. (1991) Current topic: pre-eclampsia and the placenta. Placenta, 12(4), p.301-308. Redman, C.W., Sacks, G.P. & Sargent, I.L. (1999) Preeclampsia: an excessive maternal inflammatory response to pregnancy. American Journal of Obstetrics and Gynecology, 180(0002-9378; 2), p.499-506. Regan, L. (1992) Recurrent early pregnancy failure. Curr.Opin.Obstet Gynecol, 4(2), p.220-228. Rehn, D., Reimann, H.J., von der, O.M., Schmidt, U., Schmel, A. & Hennings, G. (1987) Biorhythmic changes of plasma histamine levels in healthy volunteers. Agents Actions, 22(0065-4299; 1-2), p.24-29. Reilly, F.D. & Russell, P.T. (1977) Neurohistochemical evidence supporting an absence of adrenergic and cholinergic innervation in the human placenta and umbilical cord. The Anatomical Record, 188(3), p.277-286. Reite, O.B. (1972) Comparative physiology of histamine. Physiol Rev., 52(0031-9333; 3), p.778-819. Reynolds, L.P., Borowicz, P.P., Caton, J.S., Vonnahme, K.A., Luther, J.S., Buchanan, D.S., Hafez, S.A., Grazul-Bilska, A.T. & Redmer, D.A. (2010) Uteroplacental vascular development and placental function: an update. The International Journal of Developmental Biology, 54(1696- 3547; 0214-6282; 2-3), p.355-366. Riedel, F., Achenbach, U. & Rieger, C.H. (1989) Prematurity and maternal bronchial hyperresponsiveness. J.Perinat.Med., 17(0300-5577; 2), p.151-155. Rivera, E.S., Cricco, G.P., Engel, N.I., Fitzsimons, C.P., Martin, G.A. & Bergoc, R.M. (2000) Histamine as an autocrine growth factor: an unusual role for a widespread mediator. Seminars in Cancer Biology, 10(1), p.15-23. Roberts, J.M. & Hubel, C.A. (2009) The two stage model of preeclampsia: variations on the theme. Placenta, 30 Suppl A, p.S32-7. Roberts, J.M., Taylor, R.N., Musci, T.J., Rodgers, G.M., Hubel, C.A. & McLaughlin, M.K. (1989) Preeclampsia: an endothelial cell disorder. American Journal of Obstetrics and Gynecology, 161(5), p.1200-1204. Romero, R., Dey, S.K. & Fisher, S.J. (2014) Preterm labor: one syndrome, many causes. Science (New York, N.Y.), 345(6198), p.760-765. Rothschild, Z. & Schayer, R.W. (1958) Synthesis and metabolism of a histamine metabolite, 1-methyl- 4 (beta-aminoethyl)-imidazole. Biochim.Biophys.Acta, 30(0006-3002; 1), p.23-27. RStudio, T. (2015) RStudio: Integrated Development for R. RStudio, Inc., Boston, MA [Online]. Available at: http://www.rstudio.com/. Rukavina, D. & G, V. (2000) Role of cytokines and immune cells at the foeto-maternal interface - a workshop report. Sacks, G., Sargent, I. & Redman, C. (1999) An innate view of human pregnancy. Immunology Today, 20(0167-5699; 3), p.114-118. Sacks, G.P., Studena, K., Sargent, K. & Redman, C.W. (1998) Normal pregnancy and preeclampsia both produce inflammatory changes in peripheral blood leukocytes akin to those of sepsis. American Journal of Obstetrics and Gynecology, 179(0002-9378; 1), p.80-86. Saeed, A.I., Bhagabati, N.K., Braisted, J.C., Liang, W., Sharov, V., Howe, E.A., Li, J., Thiagarajan, M., White, J.A. & Quackenbush, J. (2006) TM4 microarray software suite. Methods in Enzymology, 411(0076-6879; 0076-6879), p.134-193. Saeed, A.I., Sharov, V., White, J., Li, J., Liang, W., Bhagabati, N., Braisted, J., Klapa, M., Currier, T., Thiagarajan, M., Sturn, A., Snuffin, M., Rezantsev, A., Popov, D., Ryltsov, A., Kostukovich, E.,

138 Borisovsky, I., Liu, Z., Vinsavich, A., Trush, V. & Quackenbush, J. (2003) TM4: a free, open- source system for microarray data management and analysis. BioTechniques, 34(0736-6205; 0736-6205; 2), p.374-378. Saito, S. & Sakai, M. (2003) Th1/Th2 balance in preeclampsia. Journal of Reproductive Immunology, 59(2), p.161-173. Saito, S., Umekage, H., Sakamoto, Y., Sakai, M., Tanebe, K., Sasaki, Y. & Morikawa, H. (1999) Increased T-helper-1-type immunity and decreased T-helper-2-type immunity in patients with preeclampsia. American Journal of Reproductive Immunology (New York, N.Y. : 1989), 41(1046- 7408; 1046-7408; 5), p.297-306. Salamonsen, L.A., Nie, G., Hannan, N.J. & Dimitriadis, E. (2009) Society for Reproductive Biology Founders' Lecture 2009. Preparing fertile soil: the importance of endometrial receptivity. Reprod.Fertil.Dev., 21(1031-3613; 1031-3613; 7), p.923-934. Sargent, I.L., Borzychowski, A.M. & Redman, C.W. (2006) Immunoregulation in normal pregnancy and pre-eclampsia: an overview. Reproductive Biomedicine Online, 13(5), p.680-686. Sargent, I.L., Germain, S.J., Sacks, G.P., Kumar, S. & Redman, C.W. (2003) Trophoblast deportation and the maternal inflammatory response in pre-eclampsia. Journal of Reproductive Immunology, 59(0165-0378; 2), p.153-160. Sasiak, K., Kierska, D., Boguslawski, M. & Maslinski, C. (1975) Phosphopyridoxal complexes with histamine and histidine. (3) The influence of presumed complex on activity of rat intestinal histaminase. Agents Actions, 5(0065-4299; 0065-4299; 1), p.25-30. Sawin, S.W., Loret, d.M.,Jr., Monzon-Bordonaba, F., Wang, C.L. & Feinberg, R.F. (1997) Hydrosalpinx fluid enhances human trophoblast viability and function in vitro: implications for embryonic implantation in assisted reproduction. Fertil.Steril., 68(0015-0282; 0015-0282; 1), p.65-71. Say, L., Chou, D., Gemmill, A., Tuncalp, O., Moller, A.B., Daniels, J., Gulmezoglu, A.M., Temmerman, M. & Alkema, L. (2014) Global causes of maternal death: a WHO systematic analysis. The Lancet.Global Health, 2(6), p.e323-33. Schayer, R.W. (1959) Catabolism of physiological quantities of histamine in vivo. Physiol Rev., 39(0031-9333; 1), p.116-126. Schayer, R.W. (1956) The metabolism of histamine in various species. Br.J.Pharmacol., 11(0007- 1188; 4), p.472-473. Schayer, R.W. (1952a) Biogenesis of histamine. J.Biol.Chem., 199(0021-9258; 1), p.245-250. Schayer, R.W. (1952b) The metabolism of ring-labeled histamine. J.Biol.Chem., 196(0021-9258; 2), p.469-475. Schayer, R.W. & Karjala, S.A. (1956) Ring N methylation; a major route of histamine metabolism. J.Biol.Chem., 221(0021-9258; 1), p.307-313. Schayer, R.W. & Reilly, M.A. (1975) Methyl derivatives of histamine; interaction with histamine metabolism. Agents Actions, 5(0065-4299; 3), p.231-235. Schena, M. (2003) Microarray Analysis. Hoboken, New Jersey: John Wiley & Sons Inc. Schena, M., Shalon, D., Davis, R.W. & Brown, P.O. (1995) Quantitative monitoring of gene expression patterns with a complementary DNA microarray. Science, 270(0036-8075; 0036- 8075; 5235), p.467-470. Schmetterer, L., Wolzt, M., Graselli, U., Findl, O., Strenn, K., Simak, S., Kastner, J., Eichler, H.G. & Singer, E.A. (1997) Nitric oxide synthase inhibition in the histamine headache model. Cephalalgia : An International Journal of Headache, 17(0333-1024; 3), p.175-182. Seita, J., Sahoo, D., Rossi, D.J., Bhattacharya, D., Serwold, T., Inlay, M.A., Ehrlich, L.I., Fathman, J.W., Dill, D.L. & Weissman, I.L. (2012) Gene Expression Commons: an open platform for absolute gene expression profiling. PloS One, 7(7), p.e40321. Sharma, S.C., Molloy, A., Walzman, M. & Bonnar, J. (1981) Levels of total ascorbic acid and histamine in the blood of women during the 3rd trimester of normal pregnancy. Int.J.Vitam.Nutr.Res., 51(0300-9831; 3), p.266-273.

139 Sharma, S.C., Sabra, A., Molloy, A. & Bonnar, J. (1984) Comparison of blood levels of histamine and total ascorbic acid in pre-eclampsia with normal pregnancy. Hum.Nutr.Clin.Nutr., 38(0263-8290; 1), p.3-9. Sharma, S.C., Walzman, M., Bonnar, J. & Molloy, A. (1985) Blood ascorbic acid and histamine levels in patients with placental bleeding. Hum.Nutr.Clin.Nutr., 39(0263-8290; 3), p.233-238. Shrivastav, P., Gill, D.S., Jeremy, J.Y., Craft, I. & Dandona, P. (1988) Follicular fluid histamine concentrations in infertile women with pelvic adhesions. Acta Obstetricia Et Gynecologica Scandinavica, 67(0001-6349; 0001-6349; 8), p.727-729. Shterental', I.S., Nikolaev, K.I., Merzhievskaia, V.M., Pikovskaia, N.B., Snegurova, V.G., Podkolodnaia, O.A., Skvortsova, I. & Pekker, M.S. (1991) Responses of the pressor and depressor regulatory systems to high sodium intake in borderline arterial hypertension. Kardiologiia, 31(10), p.47-50. Siman, C.M., Sibley, C.P., Jones, C.J., Turner, M.A. & Greenwood, S.L. (2001) The functional regeneration of syncytiotrophoblast in cultured explants of term placenta. American Journal of Physiology. Regulatory, Integrative and Comparative Physiology, 280(0363-6119; 0363-6119; 4), p.R1116-R1122. Simon, R., Lam, A., Li, M.C., Ngan, M., Menenzes, S. & Zhao, Y. (2007) Analysis of Gene Expression Data Using BRB-Array Tools. Cancer Inform., 3(1176-9351), p.11-17. Simon, R., Radmacher, M.D. & Dobbin, K. (2002) Design of studies using DNA microarrays. Genetic Epidemiology, 23(0741-0395; 1), p.21-36. Simon, R., Radmacher, M.D., Dobbin, K. & McShane, L.M. (2003) Pitfalls in the use of DNA microarray data for diagnostic and prognostic classification. J.Natl.Cancer Inst., 95(0027-8874; 1), p.14-18. Sitras, V., Fenton, C. & Acharya, G. (2015) Gene expression profile in cardiovascular disease and preeclampsia: a meta-analysis of the transcriptome based on raw data from human studies deposited in Gene Expression Omnibus. Placenta, 36(2), p.170-178. Sooranna, S.R., Lee, Y., Kim, L.U., Mohan, A.R., Bennett, P.R. & Johnson, M.R. (2004) Mechanical stretch activates type 2 cyclooxygenase via activator protein-1 transcription factor in human myometrial cells. Mol.Hum.Reprod., 10(1360-9947; 1360-9947; 2), p.109-113. Sooranna, S.R., Oteng-Ntim, E., Meah, R., Ryder, T.A. & Bajoria, R. (1999) Characterization of human placental explants: morphological, biochemical and physiological studies using first and third trimester placenta. Human Reproduction (Oxford, England), 14(2), p.536-541. Southren, A.L., Kobayashi, Y., Carmody, N.C. & Weingold, A.B. (1966a) Serial measurements of plasma diamine oxidase (DAO) during normal human pregnancy by an improved method. Evidence for the presence of a circulating DAO inhibitor. American Journal of Obstetrics and Gynecology, 95(0002-9378; 5), p.615-620. Southren, A.L., Kobayashi, Y., Sherman, C.B., Levine, L., Gordon, G. & Weingold, A.B. (1963) Plasma Diamine Oxidase in Normal and Abnormal Human Pregnancy. Surg.Forum, 14(0071- 8041), p.395-397. Southren, A.L., Kobayashi, Y., Weingold, A.B. & Carmody, N.C. (1966b) Serial plasma diamine oxidase (DAO) assays in first and second trimester complications of pregnancy. American Journal of Obstetrics and Gynecology, 96(0002-9378; 4), p.502-510. Steel, J.H., Malatos, S., Kennea, N., Edwards, A.D., Miles, L., Duggan, P., Reynolds, P.R., Feldman, R.G. & Sullivan, M.H. (2005) Bacteria and inflammatory cells in foetal membranes do not always cause preterm labor. Pediatric Research, 57(3), p.404-411. Subramanian, A., Kuehn, H., Gould, J., Tamayo, P. & Mesirov, J.P. (2007) GSEA-P: a desktop application for Gene Set Enrichment Analysis. Bioinformatics (Oxford, England), 23(1460-2059; 23), p.3251-3253. Subramanian, A., Tamayo, P., Mootha, V.K., Mukherjee, S., Ebert, B.L., Gillette, M.A., Paulovich, A., Pomeroy, S.L., Golub, T.R., Lander, E.S. & Mesirov, J.P. (2005) Gene set enrichment analysis: a knowledge-based approach for interpreting genome-wide expression profiles. Proc.Natl.Acad.Sci.U.S.A, 102(0027-8424; 43), p.15545-15550.

140 Sutherland, A., Cooper, D.W., Howie, P.W., Liston, W.A. & MacGillivray, I. (1981) The indicence of severe pre-eclampsia amongst mothers and mothers-in-law of pre-eclamptics and controls. Br.J.Obstet.Gynaecol., 88(0306-5456; 0306-5456; 8), p.785-791. Szelag, A., Merwid-Lad, A. & Trocha, M. (2002a) [Histamine receptors in the female reproductive system. Part II. The role of histamine in the placenta, histamine receptors and the uterus contractility]. Ginekol.Pol., 73(0017-0011; 7), p.636-644. Szelag, A., Merwid-Lad, A. & Trocha, M. (2002b) Histamine receptors in the female reproductive system. Part I. Role of the mast cells and histamine in female reproductive system. Ginekologia Polska, 73(7), p.627-635. Szewczyk, G., Szewczyk, A., Pyzlak, M., Klimkiewicz, J., Smiertka, W. & Szukiewicz, D. (2005) [Mast cells, histamine and development of the placental vascular network in pregnancies complicated by preeclampsia and intrauterine growth retardation]. Ginekol.Pol., 76(0017-0011; 0017-0011; 9), p.727-734. Szukiewicz, D., Maslinska, D., Stelmachow, J. & Wojtecka-Lukasik, E. (1995) Biogenetic amines in placental tissue. Relation to the contractile activity of the human uterus. Preliminary communication. Clin.Exp.Obstet.Gynecol., 22(0390-6663; 1), p.66-70. Szukiewicz, D., Szukiewicz, A., Maslinska, D., Gujski, M., Poppe, P. & Mazurek-Kantor, J. (1999) Mast cell number, histamine concentration and placental vascular response to histamine in preeclampsia. Inflammation Research : Official Journal of the European Histamine Research Society ... [Et Al.], 48 Suppl 1(1023-3830), p.S39-S40. Tamura, H., Horiike, K., Fukuda, H. & Watanabe, T. (1989) Kinetic studies on the inhibition mechanism of diamine oxidase from porcine kidney by aminoguanidine. Journal of Biochemistry, 105(2), p.299-306. Tao, T.W. & Hertig, A.T. (1965) Viability and Differentiation of Human Trophoblast in Organ Culture. The American Journal of Anatomy, 116(0002-9106; 0002-9106), p.315-327. Tasaka, K. (1991) Histamine and Blood. In: Uvnas (ed.) Histamine and Histamine Antagonists. Berlin: Springer-Verlag., p.473-510. Taylor, P.V. & Hancock, K.W. (1973) Viability of human trophoblast in vitro. The Journal of Obstetrics and Gynaecology of the British Commonwealth, 80(0022-3204; 0022-3204; 9), p.834-838. Taylor, R.N., de Groot, C.J., Cho, Y.K. & Lim, K.H. (1998) Circulating factors as markers and mediators of endothelial cell dysfunction in preeclampsia. Seminars in Reproductive Endocrinology, 16(1), p.17-31. Tetlow, L.C. & Woolley, D.E. (2003) Histamine stimulates the proliferation of human articular chondrocytes in vitro and is expressed by chondrocytes in osteoarthritic cartilage. Annals of the Rheumatic Diseases, 62(0003-4967; 10), p.991-994. Textor, J., Hardt, J. & Knuppel, S. (2011) DAGitty: a graphical tool for analyzing causal diagrams. Epidemiology (Cambridge, Mass.), 22(5), p.745. Thomas, A.G., Curtis, N. & Walker-Smith, A. (1991) Cyclical vomiting and hypertension with dermatographism and histamine release. Journal of the Royal Society of Medicine, 84(0141- 0768; 10), p.629. Thompson, K.L., Pine, P.S., Rosenzweig, B.A., Turpaz, Y. & Retief, J. (2007) Characterization of the effect of sample quality on high density oligonucleotide microarray data using progressively degraded rat liver RNA. BMC Biotechnology [Computer File], 7(1472-6750; 1472-6750), p.57. Thomson, N.F., Thornton, S. & Clark, J.F. (2000) The effects of placental extracts from normotensive and preeclamptic women on vasoconstriction and oxidative metabolism. Am.J.Obstet.Gynecol., 183(0002-9378; 1), p.206-210. Tornqvist, A. (1969) The effect of histaminase inhibition on the concentration and distribution of 14C histamine in blood during pregnancy. Acta Obstetricia Et Gynecologica Scandinavica, 48(0001- 6349; 2), p.272-284. Tranquilli, A.L., Dekker, G., Magee, L., Roberts, J., Sibai, B.M., Steyn, W., Zeeman, G.G. & Brown, M.A. (2014) The classification, diagnosis and management of the hypertensive disorders of pregnancy: A revised statement from the ISSHP. Pregnancy Hypertension, 4(2), p.97-104.

141 Tryding, N. & Willert, B. (1968) Determination of plasma diamine oxidase (histaminase) in clinical practice. A comparison between a biological method and a radiochemical micromethod. Scand.J.Clin.Lab Invest, 22(0036-5513; 1), p.29-32. Tufvesson, G. (1978) Purification and properties of human amniotic fluid diamine oxidase. Scand.J.Clin.Lab Invest, 38(0036-5513; 5), p.463-472. Tuimala, J., Korpelainen, E., Gentile, M., Kallio, A. & Hupponen, T. (2008) Chipster Manual [Online]. Available at: https://extras.csc.fi/biosciences/chipster-manual/. van der Pouw Kraan, T.C., Snijders, A., Boeije, L.C., de Groot, E.R., Alewijnse, A.E., Leurs, R. & Aarden, L.A. (1998) Histamine inhibits the production of interleukin-12 through interaction with H2 receptors. J.Clin.Invest, 102(0021-9738; 10), p.1866-1873. Vanina, L.V., Strongina, T.N. & Meshcheriakova, S.A. (1971) [The state of the blood histamine-- histaminase system in late toxemia of pregnancy]. Akush.Ginekol.(Mosk), 47(0002-3906; 4), p.25-28. Walker, D.W. & McLean, J.R. (1971) Absence of adrenergic nerves in the human placenta. Nature, 229(5283), p.344-345. Wang, J., Duncan, D., Shi, Z. & Zhang, B. (2013) WEB-based GEne SeT AnaLysis Toolkit (WebGestalt): update 2013. Nucleic Acids Research, 41(1362-4962; 0305-1048), p.W77-W83. Wantke, F., Proud, D., Siekierski, E. & Kagey-Sobotka, A. (1998) Daily variations of serum diamine oxidase and the influence of H1 and H2 blockers: a critical approach to routine diamine oxidase assessment. Inflammation Research : Official Journal of the European Histamine Research Society ... [Et Al.], 47(10), p.396-400. Watson, A.L., Palmer, M.E. & Burton, G. (1995) Human chorionic gonadotrophin release and tissue viability in placental organ culture. Hum.Reprod., 10(0268-1161; 0268-1161; 8), p.2159-2164. Weetman, A.P. (1999) The immunology of pregnancy. Thyroid : Official Journal of the American Thyroid Association, 9(1050-7256; 1050-7256; 7), p.643-646. Wegmann, T.G., Lin, H., Guilbert, L. & Mosmann, T.R. (1993) Bidirectional cytokine interactions in the maternal-foetal relationship: is successful pregnancy a TH2 phenomenon? Immunology Today, 14(0167-5699; 7), p.353-356. Weingold, A.B. & Southren, A.L. (1968) Diamine oxidase as an index of the foetoplacental unit. Clinical applications. Obstet Gynecol, 32(0029-7844; 5), p.593-606. Wen, S.W., Demissie, K. & Liu, S. (2001) Adverse outcomes in pregnancies of asthmatic women: results from a Canadian population. Annals of Epidemiology, 11(1047-2797; 1), p.7-12. WHO (2014) World Health Statistics. Geneva: World Health Organisation. Wicksell, F. (1949) Observations on Histamine and Histaminolysis in Pregnancy. Acta Physiologica Scandinavica, 17, p.395-414. Wilczynski, J.R. (2006) Immunological analogy between allograft rejection, recurrent abortion and pre- eclampsia - the same basic mechanism? Human Immunology, 67(7), p.492-511. Wit, E. & McClure, J. (2004) Statistics for Microarray: Design, Analysis and Inference.: John Wiley and Sons, Ltd. Wright, G.W. & Simon, R.M. (2003) A random variance model for detection of differential gene expression in small microarray experiments. Bioinformatics (Oxford, England), 19(1367-4803; 18), p.2448-2455. Xia, J., Fjell, C.D., Mayer, M.L., Pena, O.M., Wishart, D.S. & Hancock, R.E. (2013) INMEX--a web- based tool for integrative meta-analysis of expression data. Nucleic Acids Research, 41(1362- 4962; 0305-1048), p.W63-W70. Xia, J., Gill, E.E. & Hancock, R.E. (2015) NetworkAnalyst for statistical, visual and network-based meta-analysis of gene expression data. Nature Protocols, 10(6), p.823-844. Yang, Y.H., Dudoit, S., Luu, P., Lin, D.M., Peng, V., Ngai, J. & Speed, T.P. (2002) Normalization for cDNA microarray data: a robust composite method addressing single and multiple slide systematic variation. Nucleic Acids Research, 30(1362-4962; 4), p.e15.

142 Yang, Y.H. & Speed, T. (2002) Design issues for cDNA microarray experiments. Nature Reviews. Genetics, 3(1471-0056; 1471-0056; 8), p.579-588. Zeller, E.A. (1938) Über den enzymatischen Abbau von Histamin und Diaminen. 2. Mitteilung. Helvetica Chimica Acta, 21(1), p.880-890. Zhang, B., Schmoyer, D., Kirov, S. & Snoddy, J. (2004) GOTree Machine (GOTM): a web-based platform for interpreting sets of interesting genes using Gene Ontology hierarchies. BMC Bioinformatics, 5, p.16. Zhao, X., Ma, W., Das, S.K., Dey, S.K. & Paria, B.C. (2000) Blastocyst H(2) receptor is the target for uterine histamine in implantation in the mouse. Development, 127(0950-1991; 12), p.2643-2651. Zhou, Y., Fisher, S.J., Janatpour, M., Genbacev, O., Dejana, E., Wheelock, M. & Damsky, C.H. (1997) Human cytotrophoblasts adopt a vascular phenotype as they differentiate. A strategy for successful endovascular invasion? The Journal of Clinical Investigation, 99(0021-9738; 0021- 9738; 9), p.2139-2151.

143

Appendix

Appendix 1: Statistical Analysis results of maternal blood and urine histamine

Table A-1: Maternal Blood Histamine Levels

95% Confidence Interval for Mean

Condition N Mean Std. Deviation Std. Error Lower Bound Upper Bound Minimum Maximum

NP Trimester 1 2 64.0000 8.48528 6.00000 -12.2372 140.2372 58.00 70.00

NP Trimester 2 3 43.3333 11.54701 6.66667 14.6490 72.0177 30.00 50.00

NP Trimester 3 38 52.9108 8.80811 1.42886 50.0156 55.8059 35.00 74.30

PE 31 63.5613 10.49342 1.88467 59.7123 67.4103 49.10 94.30

Eclampsia 14 69.6714 10.03145 2.68102 63.8794 75.4634 54.20 87.30

Total 88 59.2547 11.86594 1.26491 56.7405 61.7688 30.00 94.30

Table A-2: Maternal Blood Histamine Trimester Multiple Comparisons

Dependent Variable: BloodHA Bonferroni

95% Confidence Interval

(I) Condition (J) Condition Mean Difference (I-J) Std. Error Sig. Lower Bound Upper Bound

NP Trimester 1 NP Trimester 2 20.66667 8.86125 .221 -4.8893 46.2226

NP Trimester 3 11.08921 7.04221 1.000 -9.2206 31.3990

PE .43871 7.08185 1.000 -19.9854 20.8628

Eclampsia -5.67143 7.33781 1.000 -26.8338 15.4909

NP Trimester 2 NP Trimester 1 -20.66667 8.86125 .221 -46.2226 4.8893

NP Trimester 3 -9.57746 5.82137 1.000 -26.3663 7.2114

PE -20.22796* 5.86926 .009 -37.1550 -3.3009

Eclampsia -26.33810* 6.17569 .001 -44.1488 -8.5273

NP Trimester 3 NP Trimester 1 -11.08921 7.04221 1.000 -31.3990 9.2206

NP Trimester 2 9.57746 5.82137 1.000 -7.2114 26.3663

PE -10.65050* 2.34930 .000 -17.4259 -3.8751

Eclampsia -16.76064* 3.03481 .000 -25.5131 -8.0082

PE NP Trimester 1 -.43871 7.08185 1.000 -20.8628 19.9854

NP Trimester 2 20.22796* 5.86926 .009 3.3009 37.1550

NP Trimester 3 10.65050* 2.34930 .000 3.8751 17.4259

Eclampsia -6.11014 3.12570 .540 -15.1247 2.9044

Eclampsia NP Trimester 1 5.67143 7.33781 1.000 -15.4909 26.8338

NP Trimester 2 26.33810* 6.17569 .001 8.5273 44.1488

NP Trimester 3 16.76064* 3.03481 .000 8.0082 25.5131

144 PE 6.11014 3.12570 .540 -2.9044 15.1247

*. The mean difference is significant at the 0.05 level.

Table A-3: Maternal Urine Histamine after Aminoguanidine Treatment Gestational Age Mean HA before AG Mean HA after AG Significance Non-Pregnant Mean (µg/24h) 6.6667 7.8333 p>0.05 N 3 3 Std. Deviation 5.13160 8.80814 Std. Error of Mean 2.96273 5.08538 Minimum (µg/24h) 2.33 2.50 Maximum (µg/24h) 12.33 18.00 Range (µg/24h) 10.00 15.50 Trimester 1 Mean (µg/24h) 44.1333 52.4000 p=0.137 N 5 5 Std. Deviation 18.31605 17.67201 Std. Error of Mean 8.19119 7.90316 Minimum (µg/24h) 17.67 21.00 Maximum (µg/24h) 61.00 62.00 Range (µg/24h) 43.33 41.00

Percentage (%) rise in urine histamine An analysis of maternal urine histamine (HA) levels after inhibition of Diamine oxidase (DAO) with aminoguanidine (AG) based on Bjuro et al 1964. Percentage (%) rise in urine histamine in pregnancy from non-pregnant = 562.00%. Percentage rise in maternal urine histamine after inhibition of DAO with aminoguanidine = 18.73%.

145

Appendix 2: Causal effect identification-Testable Implications

A. Adjustment for Total effect:

No adjustment is necessary to estimate the total effect of Elevated Histamine on Pre-eclampsia.

Adjustments for Direct Effect:

Minimal sufficient adjustment sets for estimating the direct effect of Elevated Histamine on Pre-eclampsia:

 Cell Signalling, Gene Regulation, Inflammation and Immune regulation, Junctional proteins regulation, Oxygen Sensing, Placental like Innervation, Placental like contractile activity, Placental growth, Tissue proliferation and morphogenesis

Instrumental and conditional Covariates:

 Defective_DAO_Activity  Increased_HDC_Activity

B. Testable implications- The model implies the following conditional independences:

1. Cell_Signaling ⊥ Defective_DAO_Activity | Elevated Histamine 2. Cell_Signaling ⊥ Gene_Regulation | Elevated Histamine 3. Cell_Signaling ⊥ Increased_HDC_Activity | Elevated Histamine 4. Cell_Signaling ⊥ Inflammation_and_Immune_regulation | Elevated Histamine 5. Cell_Signaling ⊥ Junctional_proteins_regulation | Elevated Histamine 6. Cell_Signaling ⊥ Oxygen_Sensing | Elevated Histamine 7. Cell_Signaling ⊥ Placental_Innervation | Elevated Histamine 8. Cell_Signaling ⊥ Placental_contractile_activity | Elevated Histamine 9. Cell_Signaling ⊥ Placental_growth | Elevated Histamine 10. Cell_Signaling ⊥ Tissue_proliferation_and_morphogenesis | Elevated Histamine 11. Defective_DAO_Activity ⊥ Gene_Regulation | Elevated Histamine 12. Defective_DAO_Activity ⊥ Increased_HDC_Activity 13. Defective_DAO_Activity ⊥ Inflammation_and_Immune_regulation | Elevated Histamine 14. Defective_DAO_Activity ⊥ Junctional_proteins_regulation | Elevated Histamine 15. Defective_DAO_Activity ⊥ Oxygen_Sensing | Elevated Histamine 16. Defective_DAO_Activity ⊥ Placental_Innervation | Elevated Histamine 17. Defective_DAO_Activity ⊥ Placental_contractile_activity | Elevated Histamine

146 18. Defective_DAO_Activity ⊥ Placental_growth | Elevated Histamine 19. Defective_DAO_Activity ⊥ Pre-eclampsia | Elevated Histamine 20. Defective_DAO_Activity ⊥ Tissue_proliferation_and_morphogenesis | Elevated Histamine 21. Gene_Regulation ⊥ Increased_HDC_Activity | Elevated Histamine 22. Gene_Regulation ⊥ Inflammation_and_Immune_regulation | Elevated Histamine 23. Gene_Regulation ⊥ Junctional_proteins_regulation | Elevated Histamine 24. Gene_Regulation ⊥ Oxygen_Sensing | Elevated Histamine 25. Gene_Regulation ⊥ Placental_Innervation | Elevated Histamine 26. Gene_Regulation ⊥ Placental_contractile_activity | Elevated Histamine 27. Gene_Regulation ⊥ Placental_growth | Elevated Histamine 28. Gene_Regulation ⊥ Tissue_proliferation_and_morphogenesis | Elevated Histamine 29. Increased_HDC_Activity ⊥ Inflammation_and_Immune_regulation | Elevated Histamine 30. Increased_HDC_Activity ⊥ Junctional_proteins_regulation | Elevated Histamine 31. Increased_HDC_Activity ⊥ Oxygen_Sensing | Elevated Histamine 32. Increased_HDC_Activity ⊥ Placental_Innervation | Elevated Histamine 33. Increased_HDC_Activity ⊥ Placental_contractile_activity | Elevated Histamine 34. Increased_HDC_Activity ⊥ Placental_growth | Elevated Histamine 35. Increased_HDC_Activity ⊥ Pre-eclampsia | Elevated Histamine 36. Increased_HDC_Activity ⊥ Tissue_proliferation_and_morphogenesis | Elevated Histamine 37. Inflammation_and_Immune_regulation ⊥ Junctional_proteins_regulation | Elevated Histamine 38. Inflammation_and_Immune_regulation ⊥ Oxygen_Sensing | Elevated Histamine 39. Inflammation_and_Immune_regulation ⊥ Placental_Innervation | Elevated Histamine 40. Inflammation_and_Immune_regulation ⊥ Placental_contractile_activity | Elevated Histamine 41. Inflammation_and_Immune_regulation ⊥ Placental_growth | Elevated Histamine 42. Inflammation_and_Immune_regulation ⊥ Tissue_proliferation_and_morphogenesis | Elevated Histamine 43. Junctional_proteins_regulation ⊥ Oxygen_Sensing | Elevated Histamine

147 44. Junctional_proteins_regulation ⊥ Placental_Innervation | Elevated Histamine 45. Junctional_proteins_regulation ⊥ Placental_contractile_activity | Elevated Histamine 46. Junctional_proteins_regulation ⊥ Placental_growth | Elevated Histamine 47. Junctional_proteins_regulation ⊥ Tissue_proliferation_and_morphogenesis | Elevated Histamine 48. Oxygen_Sensing ⊥ Placental_Innervation | Elevated Histamine 49. Oxygen_Sensing ⊥ Placental_contractile_activity | Elevated Histamine 50. Oxygen_Sensing ⊥ Placental_growth | Elevated Histamine 51. Oxygen_Sensing ⊥ Tissue_proliferation_and_morphogenesis | Elevated Histamine 52. Placental_Innervation ⊥ Placental_contractile_activity | Elevated Histamine 53. Placental_Innervation ⊥ Placental_growth | Elevated Histamine 54. Placental_Innervation ⊥ Tissue_proliferation_and_morphogenesis | Elevated Histamine 55. Placental_contractile_activity ⊥ Placental_growth | Elevated Histamine 56. Placental_contractile_activity ⊥ Tissue_proliferation_and_morphogenesis | Elevated Histamine 57. Placental_growth ⊥ Tissue_proliferation_and_morphogenesis | Elevated Histamine

148 Appendix 3: Scatterplot Up-regulated Genes ProbeSet Accession UGCluster Symbol 1552509_a_at NM_145273 Hs.657365 CD300LG 1552575_a_at NM_153344 Hs.485528 C6orf141 1553151_at AY079172 Hs.436360 ATP6V0D2 1553482_at NM_153040 Hs.367879 C15orf32 1553574_at NM_176891 Hs.682604 IFNE 1553789_a_at NM_058180 Hs.236572 C21orf58 1553973_a_at BC032003 Hs.334274 SPINK6 1554020_at BC010091 Hs.505202 BICD1 1554097_a_at BC021861 Hs.458096 LOC554202 1554640_at BC039306 Hs.591908 PALM2 1554663_a_at BC043499 Hs.325978 NUMA1 1554906_a_at BC029395 Hs.344400 MPHOSPH6 1554997_a_at AY151286 Hs.196384 PTGS2 1555758_a_at AF213040 Hs.84113 CDKN3 1556364_at AK057923 Hs.660881 LOC730057 1556409_a_at AF086184 Hs.103068 LOC100129932 1556423_at BE220445 Hs.525479 VASH1 1556695_a_at AK095719 Hs.457407 FLJ42709 1556698_a_at AI819722 Hs.605082 GPRIN3 1558075_at BM989131 Hs.619009 LOC339047 1558834_s_at AL832216 Hs.716732 C1orf62 1558930_at BC009533 Hs.559194 LOC728192 1559277_at AK093019 Hs.78061 FLJ35700 1560558_at BC013097 Hs.658575 C9orf80 1561171_a_at AK093450 Hs.569472 LOC727832 1561720_at BC042989 Hs.632229 RECQL5 1562102_at BC014579 Hs.460260 AKR1C1 1562337_at AK095468 Hs.531755 OR7D2 1562591_a_at AF548116 Hs.532138 OFCC1 1564231_at AK025109 Hs.478095 IFT80 1564315_at AK055534 Hs.545529 C8orf49 1564474_at AK055144 Hs.161338 LOC728723 1565716_at BE930017 Hs.513522 FUS 1566785_x_at AK025172 Hs.646586 LOC728806 1568638_a_at BQ024836 Hs.676257 IDO2 1570515_a_at BC029425 Hs.696158 FILIP1 201430_s_at W72516 Hs.519659 DPYSL3 201508_at NM_001552 Hs.462998 IGFBP4 201820_at NM_000424 Hs.433845 KRT5 201860_s_at NM_000930 Hs.491582 PLAT 202273_at NM_002609 Hs.509067 PDGFRB 202291_s_at NM_000900 Hs.365706 MGP 202363_at AF231124 Hs.643338 SPOCK1 202465_at NM_002593 Hs.202097 PCOLCE 202709_at NM_002023 Hs.519168 FMOD 202834_at NM_000029 Hs.19383 AGT

149 202936_s_at NM_000346 Hs.647409 SOX9 203083_at NM_003247 Hs.371147 THBS2 203085_s_at BC000125 Hs.645227 TGFB1 203151_at AW296788 Hs.194301 MAP1A 203153_at NM_001548 Hs.20315 IFIT1 203263_s_at AI625739 Hs.54697 ARHGEF9 203325_s_at AI130969 Hs.210283 COL5A1 203649_s_at NM_000300 Hs.466804 PLA2G2A 203886_s_at NM_001998 Hs.198862 FBLN2 204114_at NM_007361 Hs.369840 NID2 204301_at NM_014867 Hs.5333 KBTBD11 204337_at AL514445 Hs.386726 RGS4 204411_at NM_017596 Hs.169182 KIF21B 204533_at NM_001565 Hs.632586 CXCL10 204596_s_at U46768 Hs.25590 STC1 204719_at NM_007168 Hs.58351 ABCA8 204733_at NM_002774 Hs.79361 KLK6 204749_at NM_004538 Hs.21365 NAP1L3 204818_at NM_002153 Hs.162795 HSD17B2 204932_at BF433902 Hs.81791 TNFRSF11B 204945_at NM_002846 Hs.89655 PTPRN 204951_at NM_004310 Hs.654594 RHOH 205048_s_at NM_003832 Hs.512656 PSPH 205051_s_at NM_000222 Hs.479754 KIT 205112_at NM_016341 Hs.655033 PLCE1 205236_x_at NM_003102 Hs.2420 SOD3 205267_at NM_006235 Hs.654525 POU2AF1 205290_s_at NM_001200 Hs.73853 BMP2 205347_s_at NM_021992 Hs.56145 TMSB15A 205364_at NM_003500 Hs.444959 ACOX2 205404_at NM_005525 Hs.195040 HSD11B1 205447_s_at BE222201 Hs.713539 MAP3K12 205475_at NM_007281 Hs.7122 SCRG1 205481_at NM_000674 Hs.77867 ADORA1 205499_at NM_014467 Hs.306339 SRPX2 205500_at NM_001735 Hs.494997 C5 205624_at NM_001870 Hs.646 CPA3 205660_at NM_003733 Hs.118633 OASL 205680_at NM_002425 Hs.2258 MMP10 205695_at NM_006843 Hs.439023 SDS 205767_at NM_001432 Hs.115263 EREG 205782_at NM_002009 Hs.567268 FGF7 205801_s_at NM_015376 Hs.143674 RASGRP3 205870_at NM_000623 Hs.654542 BDKRB2 205890_s_at NM_006398 Hs.44532 UBD 206224_at NM_001898 Hs.123114 CST1 206227_at NM_003613 Hs.442180 CILP 206304_at NM_004997 Hs.927 MYBPH

150 206481_s_at NM_001290 Hs.714330 LDB2 206510_at AF332197 Hs.101937 SIX2 206574_s_at NM_007079 Hs.43666 PTP4A3 206618_at NM_003855 Hs.469521 IL18R1 206637_at NM_014879 Hs.2465 P2RY14 206806_at NM_004717 Hs.242947 DGKI 206924_at NM_000641 Hs.467304 IL11 206994_at NM_001899 Hs.654549 CST4 207016_s_at AB015228 Hs.643455 ALDH1A2 207191_s_at NM_005545 Hs.699822 ISLR 207276_at NM_004065 Hs.446675 CDR1 207387_s_at NM_000167 Hs.1466 GK 207510_at NM_000710 Hs.525572 BDKRB1 207889_at NM_007101 Hs.198003 SARDH 208307_at NM_005058 Hs.380450 RBMY1A1 208851_s_at AL161958 Hs.644697 THY1 209047_at AL518391 Hs.660192 AQP1 209198_s_at BC004291 Hs.32984 SYT11 209335_at AI281593 Hs.718429 DCN 209396_s_at M80927 Hs.382202 CHI3L1 209496_at BC000069 Hs.647064 RARRES2 209560_s_at U15979 Hs.533717 DLK1 209596_at AF245505 Hs.369422 MXRA5 209747_at J03241 Hs.592317 TGFB3 209914_s_at AW149405 Hs.637685 NRXN1 209955_s_at U76833 Hs.654370 FAP 210029_at M34455 Hs.840 IDO1 210119_at U73191 Hs.411299 KCNJ15 210140_at AF031824 Hs.143212 CST7 210145_at M68874 Hs.497200 PLA2G4A 210401_at U45448 Hs.41735 P2RX1 210605_s_at BC003610 Hs.3745 MFGE8 210619_s_at AF173154 Hs.75619 HYAL1 210632_s_at L35853 Hs.463412 SGCA 210687_at BC000185 Hs.503043 CPT1A 211617_at M21191 Hs.652473 ALDOAP2 211743_s_at BC005929 Hs.512633 PRG2 211958_at R73554 Hs.607212 IGFBP5 212134_at AB014538 Hs.504062 PHLDB1 212354_at BE500977 Hs.409602 SULF1 212358_at AL117468 Hs.466539 CLIP3 212624_s_at BF339445 Hs.654534 CHN1 212670_at AA479278 Hs.647061 ELN 212713_at R72286 Hs.296049 MFAP4 212736_at BE299456 Hs.401798 C16orf45 212865_s_at BF449063 Hs.409662 COL14A1 212942_s_at AB033025 Hs.459088 KIAA1199 213001_at AF007150 Hs.653262 ANGPTL2

151 213075_at AL050002 Hs.357004 OLFML2A 213249_at AU145127 Hs.433057 FBXL7 213265_at AI570199 Hs.601055 PGA3 213280_at AK000478 Hs.499659 GARNL4 213422_s_at AW888223 Hs.558570 MXRA8 213800_at X04697 Hs.363396 CFH 213820_s_at T54159 Hs.513075 STARD5 213832_at AA530995 Hs.666367 KCND3 213974_at AB033059 Hs.459162 ADAMTSL3 214022_s_at AA749101 Hs.458414 IFITM1 214079_at AK000345 Hs.272499 DHRS2 214236_at AA166684 Hs.463295 CDC27 214761_at AW149417 Hs.530930 ZNF423 214951_at AL050358 Hs.159481 SLC26A10 214978_s_at AK023365 Hs.153648 PPFIA4 215002_at BE000837 Hs.552700 LOC100132247 215184_at AK026801 Hs.237886 DAPK2 215192_at D38500 Hs.712714 PMS2L4 215388_s_at X56210 Hs.575869 CFHR1 215543_s_at AB011181 Hs.474667 LARGE 215646_s_at R94644 Hs.643801 VCAN 215856_at AK025833 Hs.287692 SIGLEC15 216176_at AK025343 Hs.675399 HCRP1 217525_at AW305097 Hs.503500 OLFML1 217590_s_at AA502609 Hs.716816 TRPA1 217897_at NM_022003 Hs.714294 FXYD6 218285_s_at NM_020139 Hs.124696 BDH2 218332_at NM_018476 Hs.334370 BEX1 218484_at NM_020142 Hs.718455 NDUFA4L2 218638_s_at NM_012445 Hs.302963 SPON2 218729_at NM_020169 Hs.478067 LXN 218820_at NM_020215 Hs.6434 C14orf132 218950_at NM_022481 Hs.25277 ARAP3 219025_at NM_020404 Hs.195727 CD248 219054_at NM_024563 Hs.13528 C5orf23 219059_s_at AL574194 Hs.655332 LYVE1 219087_at NM_017680 Hs.435655 ASPN 219230_at NM_018286 Hs.173233 TMEM100 219315_s_at NM_024600 Hs.459652 TMEM204 219359_at NM_025092 Hs.353181 ATHL1 219407_s_at NM_006059 Hs.201805 LAMC3 219478_at NM_021197 Hs.36688 WFDC1 219500_at NM_013246 Hs.502977 CLCF1 219554_at NM_016321 Hs.459284 RHCG 219602_s_at NM_022068 Hs.585839 FAM38B 219629_at NM_017911 Hs.265018 FAM118A 219689_at NM_020163 Hs.59729 SEMA3G 219700_at NM_020405 Hs.125036 PLXDC1

152 219743_at NM_012259 Hs.144287 HEY2 219837_s_at NM_018659 Hs.13872 CYTL1 219855_at NM_018159 Hs.200016 NUDT11 220092_s_at NM_018153 Hs.165859 ANTXR1 220116_at NM_021614 Hs.98280 KCNN2 220191_at NM_019617 Hs.69319 GKN1 220301_at NM_024781 Hs.280781 CCDC102B 220480_at NM_021973 Hs.388245 HAND2 220872_at NM_018547 Hs.621377 PRO2964 221305_s_at NM_019076 Hs.554822 UGT1A8 221572_s_at AF288410 Hs.631925 SLC26A6 221730_at NM_000393 Hs.445827 COL5A2 221914_at H19843 Hs.225936 SYN1 221992_at AI925734 Hs.448833 NPIPL2 222253_s_at AL117484 Hs.534980 POM121L9P 222379_at AI002715 Hs.348522 KCNE4 222450_at AL035541 Hs.517155 PMEPA1 223316_at AL136562 Hs.498720 CCDC3 223467_at AF069506 Hs.25829 RASD1 223499_at AF329841 Hs.632102 C1QTNF5 223618_at AF225426 Hs.24889 FMN2 223672_at AL136561 Hs.132121 SGIP1 223816_at AF242557 Hs.512668 SLC46A2 223987_at AF332891 Hs.432379 CHRDL2 224212_s_at AF169689 Hs.199343 PCDHA10 224339_s_at AB056476 Hs.591474 ANGPTL1 224403_at AF343661 Hs.120260 FCRL4 224976_at R37335 Hs.191911 NFIA 226069_at AA404269 Hs.524348 PRICKLE1 226304_at AA563621 Hs.534538 HSPB6 226408_at AA905942 Hs.515534 TEAD2 226474_at AA005023 Hs.528836 NLRC5 226571_s_at N38920 Hs.712625 PTPRS 226614_s_at BE856336 Hs.124299 FAM167A 226658_at AW590196 Hs.468675 PDPN 226677_at AF141339 Hs.116935 ZNF521 226766_at AB046788 Hs.13305 ROBO2 226814_at AI431730 Hs.656071 ADAMTS9 226828_s_at AL040198 Hs.472566 HEYL 226847_at BF438173 Hs.9914 FST 226971_at AI678057 Hs.521178 CCDC136 226997_at W74476 Hs.12680 ADAMTS12 227006_at AA156998 Hs.631569 PPP1R14A 227058_at AW084730 Hs.646647 C13orf33 227059_at AI651255 Hs.444329 GPC6 227099_s_at AW276078 Hs.714890 LOC387763 227154_at AL566367 Hs.212511 IGSF21 227289_at AU119437 Hs.106511 PCDH17

153 227300_at AL521682 Hs.449718 TMEM119 227475_at AI676059 Hs.591352 FOXQ1 227557_at AI127800 Hs.474251 SCARF2 227566_at AW085558 Hs.504352 NTM 227654_at AI056877 Hs.372578 FAM65C 227688_at AK022128 Hs.65366 LRCH2 227850_x_at AW084544 Hs.415791 CDC42EP5 227923_at BF439330 Hs.149035 SHANK3 227946_at AI955239 Hs.463320 OSBPL7 227984_at BE464483 Hs.371980 LOC650392 228384_s_at AI690274 Hs.238303 C10orf33 228537_at AI797248 Hs.111867 GLI2 228554_at AL137566 Hs.32405 PGR 228580_at AI828007 Hs.479119 HTRA3 228593_at AI271425 Hs.471067 LOC339483 228608_at N49852 Hs.525146 NALCN 228617_at AA142842 Hs.441975 XAF1 228875_at AI540210 Hs.126712 FAM162B 229019_at AI694320 Hs.655005 ZNF385B 229088_at BF591996 Hs.527295 ENPP1 229151_at BE673587 Hs.101307 SLC14A1 229158_at AW082836 Hs.105448 WNK4 229177_at AI823572 Hs.11782 C16orf89 229288_at BF439579 Hs.73962 EPHA7 229529_at AI827830 Hs.78061 TCF21 229678_at AA418402 Hs.380738 LOC728431 229902_at AW083785 Hs.646917 FLT4 230015_at AV729651 Hs.634380 PRCD 230109_at AI638433 Hs.594417 PDE7B 230204_at AU144114 Hs.2799 HAPLN1 230453_s_at AW188009 Hs.513870 ATP2A3 231358_at BE465760 Hs.30495 MRO 231406_at AW205664 Hs.363308 ORAI2 232001_at AW193600 Hs.590987 LOC439949 232224_at AI274095 Hs.89983 MASP1 232313_at AL122107 Hs.49599 TMEM132C 232570_s_at AL356755 Hs.173716 ADAM33 232602_at AL050348 Hs.419126 WFDC3 232645_at AW665885 Hs.259625 LOC153684 232686_at AI801574 Hs.132045 SIGLECP3 233092_s_at AL133561 Hs.648234 LOC100271840 233947_s_at U47671 Hs.567915 LOC255480 234994_at AA088177 Hs.591341 TMEM200A 235126_at N51468 Hs.552649 LQK1 235334_at AW963951 Hs.337040 ST6GALNAC3 235521_at AW137982 Hs.659337 HOXA3 235548_at BG326592 Hs.119286 APCDD1L 235588_at AA740849 Hs.99480 ESCO2

154 235874_at AL574912 Hs.98381 PRSS35 235944_at BF446673 Hs.58877 HMCN1 236044_at BF130943 Hs.40479 PPAPDC1A 236141_at AA156933 Hs.61435 NBLA00301 236738_at AW057589 Hs.710781 LOC401097 236918_s_at AW975772 Hs.591289 LRRC34 237721_s_at BE220587 Hs.666357 ASB4 238370_x_at AI252081 Hs.515329 RPL22 238542_at AA831769 Hs.656778 ULBP2 238577_s_at AA628481 Hs.271605 TSHZ2 238654_at W79425 Hs.293236 LOC147645 239077_at W81648 Hs.657569 CSGALNACT2 239286_at AI038737 Hs.116471 CDH11 239580_at BF724601 Hs.24258 GUCY1A3 239627_at BG034114 Hs.279929 TMED9 239739_at AW452218 Hs.483200 SNX24 240068_at H08345 Hs.106234 C21orf130 241221_at BE644691 Hs.505601 SEC14L3 241394_at AA213799 Hs.464224 LOC284120 241926_s_at AA296657 Hs.473819 ERG 242100_at AI076484 Hs.213137 CHSY3 242625_at AW189843 Hs.17518 RSAD2 243526_at AI968904 Hs.647083 WDR86 244413_at AW237307 Hs.560087 CLECL1 244723_at BF510430 Hs.656497 LOC100129488 336_at D38081 Hs.442530 TBXA2R 37408_at AB014609 Hs.7835 MRC2 40687_at M96789 Hs.296310 GJA4 61734_at AI797684 Hs.567550 RCN3

155 Appendix 4: Scatterplot Down-regulated Genes ProbeSet Accession UGCluster Symbol 1553243_at NM_032817 Hs.498586 ITIH5 1553333_at NM_152367 Hs.376194 C1orf161 1553352_x_at AF513360 ERVWE1 1553405_a_at NM_033225 Hs.571466 CSMD1 1553698_a_at NM_145257 Hs.715507 C1orf96 1553986_at BC023566 Hs.657750 RASEF 1554036_at BC036731 Hs.409876 ZBTB24 1554099_a_at BC032490 Hs.522672 SPIN3 1554122_a_at BC012536 Hs.132513 HSD17B12 1554195_a_at BC021680 Hs.660038 C5orf46 1554737_at AF193046 Hs.519294 FBN2 1554875_at BC029359 Hs.468349 C2orf34 1554916_a_at BC043351 Hs.535903 JRK 1555038_at BC031042 Hs.584954 EPB41L4A 1556200_a_at BC021140 Hs.587663 C10orf90 1557047_at BC032368 Hs.632575 YEATS2 1557161_at BC039503 Hs.62646 LOC100132735 1557321_a_at AA743820 Hs.468059 CAPN14 1557636_a_at BC031107 Hs.258357 C7orf57 1558212_at BC004474 Hs.416043 FLJ35024 1558451_at AK094945 Hs.435077 LOC285500 1558523_at AJ420563 Hs.443789 FAM184A 1559254_at BI826147 Hs.534828 NCRNA00162 1560932_at AK055918 Hs.562970 FLJ31356 1561514_at BC034583 Hs.579378 LOC400655 1562054_at BF377197 Hs.350673 SMEK3P 1562966_at BC017424 Hs.445885 KIAA1217 1563900_at AK055204 Hs.657974 FAM83B 1563933_a_at AK091691 Hs.672452 PLD5 1564166_s_at AK098276 Hs.406395 PRKRIP1 1568849_at BC009635 Hs.382029 C21orf135 1568882_at BC035223 Hs.317243 LRTOMT 1569112_at AW020413 Hs.654821 SLC44A5 1569464_at BC001560 Hs.172445 PPFIBP1 1569555_at BC012859 Hs.494163 GDA 1569723_a_at BC011119 Hs.461786 SPIRE2 1570394_at BC039314 Hs.435103 XRN1 201909_at NM_001008 Hs.282376 RPS4Y1 202203_s_at NM_001144 Hs.295137 AMFR 202966_at NM_014289 Hs.496593 CAPN6 203589_s_at NM_006286 Hs.379018 TFDP2 203591_s_at NM_000760 Hs.524517 CSF3R 203815_at NM_000853 Hs.268573 GSTT1 203824_at NM_004616 Hs.170563 TSPAN8 203949_at NM_000250 Hs.458272 MPO 204154_at NM_001801 Hs.442378 CDO1

156 204199_at NM_014636 Hs.648175 RALGPS1 204359_at NM_013231 Hs.533710 FLRT2 204409_s_at BC005248 Hs.461178 EIF1AY 204529_s_at AI961231 Hs.491805 TOX 204548_at NM_000349 Hs.521535 STAR 204614_at NM_002575 Hs.594481 SERPINB2 204624_at NM_000053 Hs.492280 ATP7B 204750_s_at BF196457 Hs.95612 DSC2 204844_at L12468 Hs.435765 ENPEP 205000_at NM_004660 Hs.99120 DDX3Y 205073_at NM_000775 Hs.152096 CYP2J2 205190_at NM_002670 Hs.203637 PLS1 205287_s_at NM_003222 Hs.473152 TFAP2C 205363_at NM_003986 Hs.591996 BBOX1 205543_at NM_014278 Hs.135554 HSPA4L 205916_at NM_002963 Hs.112408 S100A7 205922_at NM_004665 Hs.293130 VNN2 206239_s_at NM_003122 Hs.407856 SPINK1 206424_at NM_000783 Hs.150595 CYP26A1 206605_at NM_006025 Hs.997 P11 206698_at NM_021083 Hs.78919 XK 206700_s_at NM_004653 Hs.80358 JARID1D 206801_at NM_002521 Hs.219140 NPPB 206882_at NM_005071 Hs.515217 SLC1A6 207111_at NM_001974 Hs.2375 EMR1 207220_at NM_021071 Hs.591158 ART4 207291_at NM_024081 Hs.471695 PRRG4 207423_s_at AF029899 Hs.177984 ADAM20 207790_at NM_025168 Hs.646997 LRRC1 208607_s_at NM_030754 Hs.1955 SAA2 209521_s_at AF286598 Hs.528051 AMOT 209686_at BC001766 Hs.422181 S100B 209757_s_at BC002712 Hs.25960 MYCN 210141_s_at M13981 Hs.407506 INHA 210172_at D26121 Hs.502829 SF1 210437_at BC002351 Hs.512582 MAGEA9 210548_at U58913 Hs.169191 CCL23 210984_x_at U95089 Hs.488293 EGFR 211149_at AF000994 Hs.115277 UTY 211170_s_at AF127480 Hs.348762 PDE10A 211469_s_at U73531 Hs.34526 CXCR6 213307_at AF131790 Hs.268726 SHANK2 213816_s_at AA005141 Hs.132966 MET 213849_s_at AA974416 Hs.655213 PPP2R2B 214051_at BF677486 Hs.675540 TMSB15B 214357_at AL035295 Hs.517991 C1orf105 214397_at AI827820 Hs.25674 MBD2 214920_at R33964 Hs.120855 THSD7A

157 214983_at AL080135 Hs.433656 TTTY15 214998_at AF090100 Hs.468878 AAK1 215047_at AL080170 Hs.323858 TRIM58 215153_at AF037070 Hs.655000 NOS1AP 215491_at AI273812 Hs.437922 MYCL1 216213_at AF155113 Hs.481181 NEK1 217477_at U78581 Hs.534371 PIP5K1B 217589_at AW300309 Hs.27453 RAB40A 218182_s_at NM_021101 Hs.439060 CLDN1 218701_at NM_016027 Hs.118554 LACTB2 219263_at NM_024539 Hs.496542 RNF128 219304_s_at NM_025208 Hs.352298 PDGFD 219424_at NM_005755 Hs.501452 EBI3 219756_s_at NM_024921 Hs.267038 POF1B 219768_at NM_024626 Hs.546434 VTCN1 219814_at NM_018388 Hs.105134 MBNL3 220100_at NM_018484 Hs.220844 SLC22A11 220196_at NM_024690 Hs.432676 MUC16 220410_s_at NM_018627 Hs.522493 CAMSAP1 220528_at NM_018399 Hs.183656 VNN3 222456_s_at BF197289 Hs.525419 LIMA1 224168_at AL136742 Hs.98712 TXNDC2 224482_s_at BC006240 Hs.406788 RAB11FIP4 226665_at AI986239 Hs.655602 AHSA2 227128_s_at AI345950 Hs.23582 TACSTD2 227174_at Z98443 Hs.122125 WDR72 227241_at R79759 Hs.407152 MUC15 228212_at AL574699 Hs.29742 ISM2 228236_at AA903862 Hs.283865 C20orf54 228492_at AV681765 Hs.598540 USP9Y 229215_at AI393930 Hs.152475 ASCL2 229403_at AI572046 Hs.272011 B4GALT1 229546_at AI378035 Hs.657296 LOC653602 229599_at AA675917 Hs.390599 LOC440335 230760_at BF592062 Hs.522845 ZFY 230763_at AA905508 Hs.171130 SPATA17 230792_at BE671210 Hs.496205 FAAH2 230863_at R73030 Hs.657729 LRP2 230882_at AA129217 Hs.34969 FLJ34048 230910_s_at AI828018 Hs.69517 LY6K 231186_at AI184196 Hs.445241 FLJ43390 231240_at AI038059 Hs.202354 DIO2 231569_at N58489 Hs.98843 TMEM31 232056_at AW470178 Hs.534699 SCEL 232124_at AL117530 Hs.146346 LOC729085 232170_at AJ243672 Hs.442337 S100A7A 232388_at AB051550 Hs.461389 CNTNAP4 232452_at AI808477 Hs.23491 OR2C3

158 232771_at Z83850 Hs.209527 NRK 232849_at AI761436 Hs.204945 LOC100128988 233520_s_at AL359338 Hs.482625 CMYA5 233737_s_at AK023548 Hs.504540 LOC284561 233830_at AK023635 Hs.635164 LOC90246 234766_at AF162668 Hs.504212 OR8D2 235382_at AI246369 Hs.98288 LVRN 235942_at AI272059 Hs.568127 LOC401629 236255_at BG026457 Hs.535800 PLEKHG4B 236518_at BE208843 Hs.370555 C9orf86 236707_at AA521016 Hs.436271 DAPP1 237229_at AI268287 Hs.145717 JMJD5 237265_at BF062257 Hs.662737 C16orf73 237449_at BF447038 Hs.195922 SP8 237577_at BE465316 Hs.545311 PCNP 238263_at AW590543 Hs.642649 LOC285965 238763_at AI539118 Hs.715766 RBM20 238778_at AI244661 Hs.499159 MPP7 238790_at BE738988 Hs.659158 LOC374443 239186_at AI347139 Hs.8162 MGC39372 239243_at AA279654 Hs.434401 ZNF638 239398_at AI743156 Hs.131064 KLHL31 239430_at AA195677 Hs.546554 IGFL1 239921_at AA995791 Hs.491104 COL28A1 240091_at AI001156 Hs.464813 PSMA8 240189_at BF064226 Hs.253320 ACOXL 241111_at AI032819 Hs.364941 HSD3B1 241782_at AI932350 Hs.5025 NEBL 241881_at N54813 Hs.269151 OR2W3 241931_at AI168338 Hs.179675 XG 243541_at AI123586 Hs.55378 IL31RA 243683_at H43976 Hs.326387 MORF4L2 243709_at BG054799 Hs.649685 SLC38A9 244198_at AA885461 Hs.410810 RANBP17 244374_at N39767 Hs.515575 PLAC2 244385_at AA766126 Hs.709425 JMJD2C 244802_at AA909218 Hs.500409 GLUD1

159 Appendix 5: Differentially expressed genes among EHM and Control classes ProbeSet Symbol P-value Fold-change 1 220579_at FLJ14100 0.0003589 1.44 2 221267_s_at FAM108A1 0.0011862 1.24 3 235409_at MGA 0.001332 0.7 4 201407_s_at PPP1CB 0.0014617 0.91 5 217358_at DNAJC16 0.0017172 0.82 6 202076_at BIRC2 0.0023647 0.93 7 224151_s_at AK3 0.0026289 0.82 8 205987_at CD1C 0.002642 0.75 9 34689_at TREX1 0.0027493 1.15 10 220410_s_at CAMSAP1 0.002781 0.66 11 238790_at LOC374443 0.0028797 0.63 12 214540_at HIST1H2BO 0.0028801 0.77 13 1563933_a_at PLD5 0.0029399 0.56 14 207423_s_at ADAM20 0.0030885 0.65 15 212649_at DHX29 0.0030945 0.86 16 236329_at FLJ33996 0.0031796 1.31 17 244565_at HMX2 0.0032257 1.24 18 225210_s_at FAM103A1 0.0032481 0.82 19 1564240_at LOC100130856 0.003337 1.45 20 237229_at JMJD5 0.0034946 0.62 21 216203_at SPTLC2 0.0035257 0.75 22 215248_at GRB10 0.0035407 0.67 23 243406_at LOC440268 0.0036221 1.21 24 207959_s_at DNAH9 0.0036241 1.21 25 230015_at PRCD 0.0037443 1.7 26 243238_at PYGB 0.0037485 1.26 27 212420_at ELF1 0.0039983 0.84 28 229941_at FAM166B 0.0042202 0.9 29 216213_at NEK1 0.004344 0.59 30 217486_s_at ZDHHC17 0.0044159 0.67 31 237205_at C14orf53 0.0044451 1.15 32 238600_at JAKMIP1 0.0044965 1.31 33 240827_at FLJ45983 0.0046887 0.78 34 232318_s_at LOC121838 0.0047704 0.73 35 1553756_at C9orf70 0.0049188 1.28 36 231667_at SLC39A5 0.0049354 1.18 37 223120_at FUCA2 0.0050835 0.93 38 220947_s_at TBC1D10B 0.0052046 0.89 39 222438_at MED4 0.0052299 0.82 40 232001_at LOC439949 0.005269 1.93 41 229542_at C20orf85 0.0054122 1.24 42 217803_at GOLPH3 0.0054213 0.88 43 224827_at UBTD2 0.0055323 0.89 44 212185_x_at MT2A 0.0056176 1.1 45 228621_at HFE2 0.0058128 0.87

160 46 205500_at C5 0.0062228 1.59 47 204638_at ACP5 0.0062257 1.34 48 203250_at RBM16 0.0062952 0.93 49 224069_x_at P2RX2 0.0063464 1.3 50 202371_at TCEAL4 0.0064324 0.81 51 218171_at VPS4B 0.0065387 0.85 52 236918_s_at LRRC34 0.0065704 1.6 53 205553_s_at CSRP3 0.0065788 1.14 54 1566363_at DNTT 0.0067486 0.78 55 229860_x_at C4orf48 0.0069419 1.26 56 203167_at TIMP2 0.0071573 1.26 57 218529_at CD320 0.0071899 1.29 58 204617_s_at ACD 0.0071951 1.21 59 244119_at LOC283483 0.0073746 0.85 60 223579_s_at APOB 0.0073762 1.12 61 228154_at C19orf44 0.0074435 0.8 62 1561232_at LOC100270680 0.007454 1.11 63 230012_at C17orf44 0.007469 1.41 64 209300_s_at NECAP1 0.0076702 0.87 65 46142_at LMF1 0.007714 1.42 66 1558795_at LOC728052 0.0079144 0.87 67 222742_s_at RABL5 0.0081637 1.22 68 238644_at MYSM1 0.0082074 0.74 69 239585_at KAT2B 0.0084524 0.84 70 223430_at SIK2 0.0084569 1.16 71 208308_s_at GPI 0.008541 1.13 72 225603_s_at C8orf83 0.0086517 0.83 73 1555343_at MEGF10 0.0088899 1.29 74 209523_at TAF2 0.0089522 0.86 75 204548_at STAR 0.0089635 0.66 76 215452_x_at SUMO4 0.0090376 0.89 77 240784_at C7orf52 0.0090543 0.86 78 1554801_at C5orf40 0.0094134 0.88 79 207403_at IRS4 0.0095524 1.24 80 203593_at CD2AP 0.0095917 0.87 81 236283_x_at LOC646214 0.0096327 0.87 82 206875_s_at SLK 0.0098347 0.87 83 1559682_at TRIM16L 0.0099596 1.27 84 221336_at ATOH1 0.0100868 0.9 85 221220_s_at SCYL2 0.010151 0.68 86 233681_at KRTAP3-3 0.0106159 0.87 87 203614_at UTP14C 0.0112371 0.94 88 206355_at GNAL 0.0113893 1.39 89 221005_s_at PTDSS2 0.0115387 1.18 90 210033_s_at SPAG6 0.0115842 0.7 91 1570394_at XRN1 0.0115887 0.56 92 203686_at MPG 0.0116844 1.17 93 220022_at ZNF334 0.0119879 0.78

161 94 204992_s_at PFN2 0.012005 0.91 95 201348_at GPX3 0.0122545 0.78 96 220971_at IL25 0.0122726 0.85 97 216551_x_at PLCG1 0.0122809 1.26 98 235261_at UNC45B 0.0123232 1.25 99 230993_s_at C6orf118 0.0123761 0.91 100 234617_at OR52D1 0.0126402 0.85 101 210406_s_at RAB6C 0.0127851 0.95 102 1564974_at KRTAP8-1 0.0129876 1.21 103 206842_at KCND1 0.0130475 0.86 104 200661_at CTSA 0.0130818 1.19 105 236033_at ASB12 0.0130968 0.83 106 234226_at OPN4 0.0131863 0.9 107 237056_at INSC 0.0132595 1.4 108 223816_at SLC46A2 0.0133179 1.8 109 1563685_at LOC285422 0.0133496 0.8 110 222705_s_at SLC25A15 0.0134727 0.72 111 227120_at FOXP4 0.0134956 1.33 112 216380_x_at RPS28P6 0.0135994 1.13 113 233353_at FER1L5 0.013662 1.28 114 243683_at MORF4L2 0.0137011 0.46 115 1557296_at FLJ12825 0.0139896 1.27 116 244132_x_at ZNF518A 0.0140383 0.8 117 1558622_a_at ZNF548 0.0140483 1.23 118 203982_s_at ABCD4 0.0141782 1.19 119 238978_at LMBRD2 0.0143239 0.7 120 225193_at KIAA1967 0.0143674 0.75 121 210057_at SMG1 0.0144536 0.71 122 238833_at LOC729088 0.0144885 1.36 123 231791_at ASAH2B 0.0146016 0.68 124 202600_s_at NRIP1 0.0149217 0.71 125 212682_s_at LMF2 0.015119 1.19 126 1554882_at ERCC8 0.0152081 0.71 127 241736_at FBXW2 0.0152111 0.74 128 206427_s_at MLANA 0.0152622 1.16 129 1553106_at C5orf24 0.0153748 0.75 130 227946_at OSBPL7 0.0153901 1.76 131 1564660_at LOC100131864 0.0154189 1.09 132 213303_x_at ZBTB7A 0.0154385 1.13 133 215244_at DGCR5 0.0157839 1.28 134 209112_at CDKN1B 0.0158879 0.88 135 209752_at REG1A 0.0159607 1.22 136 229458_s_at GALK1 0.0161022 1.24 137 208551_at HIST1H4G 0.0161741 0.85 138 227617_at TMEM201 0.0161898 1.14 139 232810_at AIG1 0.016286 0.73 140 216088_s_at PSMA7 0.0163512 1.11 141 206505_at UGT2B4 0.0163536 0.9

162 142 222259_s_at SPO11 0.0164153 1.16 143 225024_at RPRD1B 0.0167583 0.88 144 216921_s_at KRT35 0.0168356 1.15 145 224252_s_at FXYD5 0.016973 1.19 146 205775_at FAM50B 0.0170546 1.43 147 214432_at ATP1A3 0.0171428 1.19 148 214903_at SYT2 0.0171937 0.9 149 1567060_at OR8G1 0.0173169 0.75 150 238577_s_at TSHZ2 0.0173521 2.59 151 237619_at C6orf146 0.01744 1.12 152 209705_at MTF2 0.0175313 0.87 153 216699_s_at KLK1 0.0177079 1.2 154 228026_at RP5-1000E10.4 0.0177405 0.84 155 32209_at FAM89B 0.0177597 1.2 156 240521_at NCRNA00153 0.0178521 1.4 157 232486_at LRFN1 0.0180814 1.14 158 215959_at PPFIBP2 0.0182769 0.78 159 241256_at LOC100131283 0.0183674 1.21 160 206424_at CYP26A1 0.0184207 0.6 161 1561384_a_at LOC284661 0.0185284 1.17 162 223273_at C14orf142 0.0186446 0.8 163 205140_at FPGT 0.0189011 0.84 164 233344_x_at KIAA1875 0.0191119 0.86 165 1564435_a_at KRT72 0.0191456 1.18 166 223780_s_at MED13 0.0192048 0.68 167 214161_at OSGIN2 0.0192178 0.76 168 241599_at LSM11 0.0192427 1.5 169 1562209_at WDR21B 0.0192807 0.75 170 226617_at ARL5A 0.0196498 0.87 171 227637_at TFCP2 0.0196867 0.76 172 219695_at SMPD3 0.0200129 0.88 173 208745_at ATP5L 0.0200659 0.88 174 231783_at CHRM1 0.0202329 0.8 175 225028_at LOC550643 0.0203 0.82 176 214995_s_at APOBEC3F 0.0203374 1.25 177 204382_at NAT9 0.0204292 1.11 178 227005_at RPP14 0.0206063 0.87 179 226390_at STARD4 0.0206713 0.9 180 1557146_a_at LOC146336 0.0207754 0.87 181 200943_at HMGN1 0.0208001 0.9 182 219273_at CCNK 0.0208133 0.73 183 1553246_a_at SGK196 0.0210413 0.77 184 220217_x_at SPANXC 0.0211315 1.13 185 219443_at TASP1 0.0213244 0.8 186 242685_at GTPBP8 0.0214123 0.77 187 209361_s_at PCBP4 0.0215662 1.39 188 217269_s_at PRSS7 0.0215683 1.34 189 230418_s_at GALNTL1 0.021664 1.21

163 190 206080_at PLCH2 0.021834 1.15 191 206358_at PRM1 0.0219524 1.19 192 206872_at SLC17A1 0.0220032 1.08 193 223754_at MGC13057 0.0220641 0.72 194 215790_at AJAP1 0.0220801 1.31 195 1552751_a_at CIB3 0.0221359 0.82 196 1560932_at FLJ31356 0.0222339 0.66 197 232039_at KIAA1383 0.0222371 0.85 198 212440_at SNRNP27 0.0224666 0.92 199 223752_at CFC1 0.02253 1.17 200 220223_at ATAD5 0.0227945 1.27 201 223506_at ZC3H8 0.022919 0.87 202 238163_at LOC100133229 0.0231235 1.18 203 1566524_a_at hCG_1986447 0.0231327 0.68 204 206605_at P11 0.023154 0.64 205 208093_s_at NDEL1 0.0237543 0.95 206 206967_at CCNT1 0.0239101 0.78 207 205948_at PTPRT 0.0239316 1.26 208 1552602_at CACNG5 0.0240234 0.78 209 209323_at PRKRIR 0.0240316 0.92 210 232291_at MIRHG1 0.0240519 0.85 211 205765_at CYP3A5 0.0241927 0.9 212 244672_at WDR1 0.0242403 0.85 213 203413_at NELL2 0.0242406 0.83 214 1555330_at GCLC 0.0243718 0.76 215 1564166_s_at PRKRIP1 0.024386 0.65 216 223330_s_at SUGT1 0.0245407 0.82 217 206151_x_at ELA3B 0.024579 1.14 218 205350_at CRABP1 0.0246988 1.42 219 226268_at RAB21 0.0247677 0.87 220 213838_at NOL7 0.024782 0.83 221 233229_at SCFD1 0.0251655 0.7 222 237449_at SP8 0.0251984 0.65 223 228139_at RIPK3 0.0252366 1.32 224 212844_at RRP1B 0.025475 0.87 225 238586_at LOC731489 0.025525 1.11 226 210454_s_at KCNJ6 0.0255429 1.37 227 208548_at IFNA6 0.0255546 1.17 228 220627_at CST8 0.0258015 1.22 229 236514_at ACOT8 0.0258242 0.8 230 238893_at LOC338758 0.0258614 1.49 231 209593_s_at TOR1B 0.0258656 1.07 232 243362_s_at LOC641518 0.0258671 1.41 233 219171_s_at ZNF236 0.025871 0.78 234 221641_s_at ACOT9 0.0261025 1.07 235 238135_at AGTRAP 0.0261472 0.8 236 226167_at SYT7 0.0261505 1.36 237 205475_at SCRG1 0.0263258 2

164 238 34225_at WHSC2 0.026431 0.84 239 1563581_at LOC285456 0.0264972 1.32 240 1553314_a_at KIF19 0.0265424 0.85 241 220224_at HAO1 0.0265452 0.86 242 226800_at EFCAB7 0.0265954 0.87 243 1555238_at PTH2 0.0266333 1.13 244 229566_at LOC645638 0.0266682 0.73 245 216607_s_at CYP51A1 0.0267076 1.33 246 209937_at TM4SF4 0.0267185 1.23 247 241779_at MTX3 0.0268559 0.77 248 238016_s_at LOC100129105 0.0268726 1.1 249 201574_at ETF1 0.0270324 0.91 250 232907_at UBR4 0.0270394 0.69 251 1562966_at KIAA1217 0.027046 0.6 252 229094_at LOC401431 0.0270462 1.22 253 226962_at ZBTB41 0.0270543 0.82 254 227365_at ATCAY 0.0273438 0.93 255 230763_at SPATA17 0.0273701 0.61 256 231731_at OTX2 0.0274355 0.88 257 218760_at COQ6 0.0276837 1.11 258 205364_at ACOX2 0.0277763 1.61 259 225235_at TSPAN17 0.0277944 1.17 260 1554339_a_at COG3 0.0279185 0.72 261 225580_at MRPL50 0.0279466 0.85 262 208440_at C3orf27 0.0280131 1.15 263 208578_at SCN10A 0.0280748 1.2 264 218174_s_at C10orf57 0.0281064 0.84 265 227358_at ZBTB46 0.0281615 1.24 266 235310_at GCET2 0.028259 0.79 267 1558027_s_at PRKAB2 0.0282756 0.88 268 237721_s_at ASB4 0.0283536 1.58 269 205911_at PTH1R 0.0283936 1.28 270 234979_at BCDIN3D 0.0284326 0.81 271 233532_x_at IFT52 0.0285195 1.19 272 208040_s_at MYBPC3 0.0287075 0.89 273 220235_s_at C1orf103 0.0287579 0.77 274 239492_at SEC14L4 0.0287998 1.4 275 230839_at PRMT8 0.0289504 1.16 276 209668_x_at CES2 0.0292073 1.13 277 204951_at RHOH 0.0292183 1.62 278 211646_at LOC100126583 0.0295025 0.88 279 203924_at GSTA2 0.0295472 1.3 280 218297_at C10orf97 0.0295589 0.84 281 208953_at LARP5 0.0296576 0.84 282 214411_x_at CTRB2 0.0297295 0.79 283 207844_at IL13 0.0297325 0.86 284 1553193_at ZNF441 0.0303803 0.73 285 210753_s_at EPHB1 0.0303997 1.47

165 286 229422_at NRD1 0.0304688 0.71 287 200627_at PTGES3 0.030551 0.93 288 231765_at ZFYVE20 0.030568 1.25 289 242284_at LOC199899 0.0305789 1.21 290 203965_at USP20 0.0305844 1.14 291 219916_s_at RNF39 0.0306563 1.32 292 212339_at EPB41L1 0.0307202 1.18 293 206578_at NKX2-5 0.0307973 1.21 294 216460_at FLJ00049 0.030816 1.23 295 210873_x_at APOBEC3A 0.0308197 1.11 296 1559561_at FBXO18 0.0308873 1.18 297 206411_s_at ABL2 0.0309285 1.36 298 1564000_at ANKRD31 0.031043 0.84 299 238061_at LGI3 0.0310584 0.83 300 235588_at ESCO2 0.0314097 1.7 301 210556_at NFATC3 0.0317492 1.39 302 242441_at LOC646548 0.0318012 0.88 303 221966_at GPR137 0.0318496 0.86 304 227720_at ANKRD13B 0.032017 1.33 305 244175_at RP3-439I14.1 0.0320294 1.11 306 205271_s_at CCRK 0.0320322 0.84 307 244198_at RANBP17 0.0320327 0.6 308 1556266_a_at LOC400831 0.0320601 1.34 309 243916_x_at UBLCP1 0.0321357 0.89 310 227601_at METTL14 0.0323088 0.88 311 230141_at ARID4A 0.0323968 0.82 312 232570_s_at ADAM33 0.0324026 1.7 313 1553191_at DST 0.0326154 0.71 314 208259_x_at IFNA7 0.0326415 0.87 315 215112_x_at MCF2L2 0.0326861 1.11 316 218299_at C11orf24 0.0327883 1.13 317 220077_at CCDC134 0.0328582 1.14 318 234025_at LOC100129890 0.0329924 1.28 319 223928_s_at GUCA1C 0.0331593 0.85 320 235866_at C9orf85 0.033266 0.72 321 228782_at SCGB3A2 0.0333221 0.88 322 215324_at SEMA3D 0.0333565 0.84 323 228001_at TMEM50B 0.0334288 1.2 324 201822_at TIMM17A 0.0334318 0.74 325 243268_at PATE2 0.0337116 0.85 326 222792_s_at CCDC59 0.0337128 1.16 327 230136_at LOC400099 0.0337718 0.8 328 219833_s_at EFHC1 0.0338373 1.24 329 242883_at OTOS 0.0338653 0.91 330 217620_s_at PIK3CB 0.0338695 0.7 331 1556133_s_at LOC100169752 0.0338836 1.12 332 204501_at NOV 0.0339183 1.18 333 237120_at KRT77 0.0339548 1.05

166 334 233171_at GRIN3A 0.0340149 0.84 335 224846_at SHKBP1 0.0340186 1.16 336 213322_at C6orf130 0.0340868 0.89 337 224110_at LOC100133319 0.0342204 1.49 338 242210_at ZNF24 0.034362 0.78 339 236421_at ANKRD45 0.0343994 0.86 340 239198_at EZH1 0.0345462 0.75 341 237513_at TRYX3 0.0346088 0.9 342 205044_at GABRP 0.0346971 0.81 343 1567238_at OR2L2 0.0349928 0.77 344 1553822_at RTP1 0.035195 1.21 345 217515_s_at CACNA1S 0.0352024 0.87 346 212160_at XPOT 0.0352788 0.89 347 1569436_at LOC400128 0.0353156 1.38 348 204030_s_at SCHIP1 0.0355554 0.91 349 233913_at WFDC10A 0.035772 0.78 350 205029_s_at FABP7 0.0359774 0.8 351 235707_at LOC221710 0.0360783 0.74 352 1552990_at LOC100133695 0.0363598 0.81 353 239143_x_at RNF138 0.0363651 0.69 354 218861_at RNF25 0.0367013 1.15 355 215491_at MYCL1 0.0368251 0.62 356 214554_at HIST1H2AL 0.0368598 1.22 357 216937_s_at RS1 0.0369858 0.76 358 219101_x_at ABHD8 0.0370082 0.9 359 206612_at CACNG1 0.0370709 0.85 360 200025_s_at RPL27 0.0370973 0.96 361 243927_x_at KIAA1429 0.0371189 0.86 362 1558584_at UBL4B 0.0371305 0.85 363 1563696_at HSD17B4 0.0375333 0.83 364 224436_s_at NIPSNAP3A 0.0376867 0.81 365 1556472_s_at SCML4 0.0377806 1.13 366 222949_at NXF3 0.0378359 0.86 367 220186_s_at PCDH24 0.0379349 1.14 368 222825_at OTUD6B 0.0379432 0.84 369 231403_at TRIO 0.0380465 0.81 370 1570006_at LOC400958 0.0380896 1.33 371 204936_at MAP4K2 0.0381501 1.19 372 237493_at IL22RA2 0.0382551 1.16 373 231719_at IFRG15 0.0383874 0.67 374 224311_s_at CAB39 0.0384198 0.86 375 205984_at CRHBP 0.0385009 1.31 376 219632_s_at TRPV1 0.0385879 1.25 377 224242_at GALP 0.0386116 1.12 378 223030_at TRAF7 0.0386461 0.84 379 218164_at SPATA20 0.0387422 1.31 380 217378_x_at LOC100130100 0.0389593 1.11 381 226654_at MUC12 0.038983 1.34

167 382 225949_at NRBP2 0.0389924 1.31 383 217731_s_at ITM2B 0.0390139 0.91 384 207694_at POU3F4 0.0392589 1.14 385 211339_s_at ITK 0.0392954 0.87 386 223740_at C6orf59 0.0393114 1.15 387 1562388_at LOC285819 0.0393585 1.11 388 219345_at BOLA1 0.0393812 1.23 389 224678_at KIAA1219 0.0394098 0.88 390 220116_at KCNN2 0.0394641 1.56 391 202941_at NDUFV2 0.0395118 1.07 392 1560147_at LOC100131176 0.0395294 1.19 393 203749_s_at RARA 0.0395419 1.3 394 235469_at FAM133B 0.0395577 0.76 395 1557571_at VPS13D 0.0397487 1.18 396 230341_x_at ADAMTS10 0.0397793 1.4 397 1559627_at LOC285941 0.0398172 1.12 398 1556200_a_at C10orf90 0.0398429 0.59 399 203253_s_at HISPPD1 0.0398832 0.87 400 213589_s_at B3GNTL1 0.0399597 1.44 401 205716_at SLC25A40 0.0399974 0.81 402 205839_s_at BZRAP1 0.0400035 1.42 403 210720_s_at NECAB3 0.0400495 1.28 404 211147_s_at P2RX6 0.0403632 0.77 405 215503_at SPINT3 0.0403652 0.88 406 1560806_at LOC150527 0.0405062 1.29 407 207849_at IL2 0.0405218 1.11 408 220972_s_at KRTAP9-9 0.0406294 0.92 409 210175_at C2orf3 0.0406319 0.69 410 204611_s_at PPP2R5B 0.0406393 1.29 411 232373_at NOXA1 0.0406835 1.08 412 239243_at ZNF638 0.0407022 0.64 413 211116_at SLC9A2 0.040738 1.17 414 231068_at SLC47A2 0.0407413 1.09 415 208479_at KCNA1 0.0408819 1.24 416 237288_at TGM7 0.0409264 0.87 417 223398_at C9orf89 0.0409987 1.16 418 209497_s_at RBM4B 0.0410305 0.93 419 1570128_at DDX19A 0.0411041 0.84 420 244694_at IGLON5 0.0411146 1.45 421 222040_at LOC728844 0.0411901 0.84 422 225607_at CCDC43 0.0412916 0.93 423 219359_at ATHL1 0.0412942 1.6 424 1558292_s_at PIGW 0.0413222 0.83 425 232143_at DNM1P41 0.0413762 1.26 426 213526_s_at LIN37 0.0413856 1.11 427 225844_at POLE4 0.0413904 1.15 428 225455_at TADA1L 0.0414091 0.85 429 238149_at ZNF818P 0.0414375 0.73

168 430 207557_s_at RYR2 0.0415176 1.48 431 221240_s_at B3GNT4 0.0415369 1.23 432 1564386_at TXNDC8 0.041683 0.85 433 241955_at HECTD1 0.0418456 0.77 434 219369_s_at OTUB2 0.0420378 0.75 435 225277_at SLC39A13 0.042045 1.16 436 212589_at RRAS2 0.0420794 0.86 437 220573_at KLK14 0.0422841 0.86 438 239802_at SAP30L 0.0423883 0.75 439 221392_at TAS2R4 0.0423904 0.76 440 1555240_s_at GNG12 0.042565 0.78 441 218103_at FTSJ3 0.0425963 1.09 442 232366_at KIAA0232 0.0426389 0.78 443 205720_at POMC 0.042663 0.85 444 207021_at ZPBP 0.0428135 1.1 445 224553_s_at TNFRSF18 0.0429597 0.81 446 209151_x_at TCF3 0.0429873 1.24 447 219095_at JMJD7- 0.0431288 1.17 PLA2G4B 448 243311_at DEFB132 0.0431385 0.9 449 219730_at MED18 0.0432211 1.2 450 1563137_at MGAT5B 0.0432279 0.87 451 1554875_at C2orf34 0.0432354 0.6 452 214226_at POL3S 0.0434023 1.22 453 216176_at HCRP1 0.0434029 1.77 454 1554916_a_at JRK 0.043474 0.67 455 211152_s_at HTRA2 0.0437694 1.09 456 225076_s_at ZNFX1 0.0437904 1.21 457 226325_at ADSSL1 0.0438138 1.25 458 241963_at ZNF704 0.0439061 1.33 459 229690_at FAM109A 0.0439114 1.15 460 205447_s_at MAP3K12 0.0439787 1.91 461 218458_at GMCL1 0.0440709 0.81 462 207830_s_at PPP1R8 0.0441136 0.95 463 222759_at SUV420H1 0.0443464 0.82 464 203765_at GCA 0.0444953 0.85 465 237021_at LOC144486 0.0445005 1.15 466 207972_at GLRA1 0.044605 1.17 467 226493_at KCTD18 0.0447031 0.71 468 235619_at LOC285986 0.0447732 1.33 469 202957_at HCLS1 0.0448217 0.9 470 231197_at PPP1R3F 0.0449499 0.8 471 228384_s_at C10orf33 0.0451921 1.53 472 215567_at FCF1 0.0453041 0.88 473 208244_at BMP3 0.0454621 1.15 474 222817_at HSD3B7 0.0454939 1.15 475 201467_s_at NQO1 0.0455492 1.4 476 230076_at PITPNM3 0.0455623 1.23

169 477 208843_s_at GORASP2 0.0455965 0.83 478 239894_at LOC100128511 0.0456483 0.82 479 1561169_at LOC727818 0.0456823 1.2 480 1555947_at FAM120A 0.0457091 0.71 481 1553267_a_at CNOT6L 0.0457881 0.68 482 234903_at OR2B3 0.0459378 0.89 483 213866_at SAMD14 0.0459837 1.43 484 219011_at PLEKHA4 0.0460468 1.41 485 1562591_a_at OFCC1 0.0461233 1.62 486 1565728_at LOC284630 0.0461407 1.18 487 206915_at NKX2-2 0.0462666 1.18 488 228880_at LOC339984 0.0465128 0.8 489 220007_at METTL8 0.0465492 0.76 490 223823_at KCNMB2 0.0465625 0.76 491 218513_at C4orf43 0.0466076 0.84 492 1553525_at NLRP13 0.0466533 1.4 493 212713_at MFAP4 0.0467887 2.8 494 1557864_x_at LOC100129266 0.0467995 1.31 495 229451_at GALNT9 0.0468526 1.11 496 1564785_at LOC196913 0.0468907 1.11 497 211170_s_at PDE10A 0.0469142 0.55 498 242200_at ADAMTSL5 0.0470495 1.27 499 218745_x_at TMEM161A 0.047207 1.21 500 1569974_x_at SEP13 0.0472981 1.14 501 212391_x_at LOC100130107 0.0473749 0.97 502 202412_s_at USP1 0.0474036 0.78 503 229399_at C10orf118 0.0475045 0.81 504 236055_at DQX1 0.0475116 0.67 505 210380_s_at CACNA1G 0.0475193 1.29 506 220347_at SMG6 0.0475337 1.2 507 203983_at TSNAX 0.047564 0.94 508 1553558_at TAS2R41 0.047615 1.23 509 219751_at SETD6 0.0476513 0.79 510 1554309_at EIF4G3 0.0478381 0.74 511 242427_at WAC 0.0478621 0.88 512 236675_at RPA1 0.0478876 0.71 513 222220_s_at TSNAXIP1 0.0478912 1.25 514 213200_at SYP 0.0479142 1.4 515 1569468_at ZNF732 0.0479338 0.84 516 231070_at IYD 0.0479356 1.13 517 206020_at SOCS6 0.0479646 0.88 518 1559252_a_at C20orf29 0.0479771 1.31 519 227325_at LOC255783 0.0480529 1.23 520 225754_at AP1G1 0.0481797 0.88 521 230759_at SNX14 0.0482163 0.91 522 218677_at S100A14 0.0483573 0.72 523 218411_s_at MBIP 0.0483993 0.89 524 235129_at PPP1R1A 0.0484194 1.31

170 525 221456_at TAS2R3 0.0488843 1.12 526 210119_at KCNJ15 0.0489331 1.95 527 209064_x_at PAIP1 0.0489473 0.83 528 213611_at AQP5 0.0489659 0.86 529 1555757_at C7orf34 0.0489967 0.9 530 1552594_at TMEM190 0.0490474 1.23 531 205014_at FGFBP1 0.04912 0.8 532 204873_at PEX1 0.049123 0.8 533 225883_at ATG16L2 0.0492266 1.33 534 1554794_a_at UBE3C 0.0492977 1.27 535 237002_at NCDN 0.0493018 1.16 536 214237_x_at LOC100133105 0.0493116 1.21 537 1568690_a_at LOC728853 0.0495655 0.88 538 239594_at LOC145837 0.0496205 1.21 539 202393_s_at KLF10 0.0498015 0.91 540 226346_at MEX3A 0.0499647 1.42

171 Appendix 6: Up-regulated HSPE Genes Gene Symbol Gene Name EntrezGene Ensembl ACTR3 ARP3 actin-related protein 3 homolog (yeast) 10096 ENSG00000115091 ADAR adenosine deaminase, RNA-specific 103 ENSG00000160710 AFF1 AF4/FMR2 family, member 1 4299 ENSG00000172493 AIDA axin interactor, dorsalization associated 64853 ENSG00000186063 ANXA3 annexin A3 306 ENSG00000138772 ANXA8L2 annexin A8-like 2 244 ENSG00000186807 APP amyloid beta (A4) precursor protein 351 ENSG00000142192 ARF1 ADP-ribosylation factor 1 375 ENSG00000143761 ARPC4 actin related protein 2/3 complex, subunit 4, 20kDa 10093 ENSG00000241553 ATP1B3 ATPase, Na+/K+ transporting, beta 3 polypeptide 483 ENSG00000069849 ATP2A2 ATPase, Ca++ transporting, cardiac muscle, slow twitch 488 ENSG00000174437 2 ATP6V1D ATPase, H+ transporting, lysosomal 34kDa, V1 subunit D 51382 ENSG00000100554 BZW1 basic leucine zipper and W2 domains 1 9689 ENSG00000082153 C15orf48 chromosome 15 open reading frame 48 84419 ENSG00000166920 CALD1 caldesmon 1 800 ENSG00000122786 CCDC6 coiled-coil domain containing 6 8030 ENSG00000108091 CCND2 cyclin D2 894 ENSG00000118971 CCT3 chaperonin containing TCP1, subunit 3 (gamma) 7203 ENSG00000163468 CD164 CD164 molecule, sialomucin 8763 ENSG00000135535 CDC42 cell division cycle 42 (GTP binding protein, 25kDa) 998 ENSG00000070831 CDKN1A cyclin-dependent kinase inhibitor 1A (p21, Cip1) 1026 ENSG00000124762 CKAP4 cytoskeleton-associated protein 4 10970 ENSG00000136026 CLDND1 claudin domain containing 1 56650 ENSG00000080822 CMTM6 CKLF-like MARVEL transmembrane domain containing 6 54918 ENSG00000091317 COX7B cytochrome c oxidase subunit VIIb 1349 ENSG00000131174 CYR61 cysteine-rich, angiogenic inducer, 61 3491 ENSG00000142871 DEK DEK oncogene 7913 ENSG00000124795 DLG5 discs, large homolog 5 (Drosophila) 9231 ENSG00000151208 DNAJB6 DnaJ (Hsp40) homolog, subfamily B, member 6 10049 ENSG00000105993 DYNLRB1 dynein, light chain, roadblock-type 1 83658 ENSG00000125971 EGR1 early growth response 1 1958 ENSG00000120738 ERGIC1 endoplasmic reticulum-golgi intermediate compartment 57222 ENSG00000113719 (ERGIC) 1 ERP29 endoplasmic reticulum protein 29 10961 ENSG00000089248 ETV5 ets variant 5 2119 ENSG00000244405 FCER1G Fc fragment of IgE, high affinity I, receptor for; gamma 2207 ENSG00000158869 polypeptide FDFT1 farnesyl-diphosphate farnesyltransferase 1 2222 ENSG00000079459 FLNB filamin B, beta 2317 ENSG00000136068 GABARAP GABA(A) receptor-associated protein 11337 ENSG00000170296 GJA1 gap junction protein, alpha 1, 43kDa 2697 ENSG00000152661 H2AFY H2A histone family, member Y 9555 ENSG00000113648 H3F3B H3 histone, family 3B (H3.3B) 3021 ENSG00000132475 HDGF hepatoma-derived growth factor 3068 ENSG00000143321 HMGB3 high mobility group box 3 3149 ENSG00000029993

172 HN1 hematological and neurological expressed 1 51155 ENSG00000189159 HNRNPA3 heterogeneous nuclear ribonucleoprotein A3 220988 ENSG00000170144 HNRNPH3 heterogeneous nuclear ribonucleoprotein H3 (2H9) 3189 ENSG00000096746 HNRNPM heterogeneous nuclear ribonucleoprotein M 4670 ENSG00000099783 HNRPDL heterogeneous nuclear ribonucleoprotein D-like 9987 ENSG00000152795 HSPA1A heat shock 70kDa protein 1A 3303 ENSG00000204389 IER2 immediate early response 2 9592 ENSG00000160888 IL24 interleukin 24 11009 ENSG00000162892 INHBC inhibin, beta C 3626 ENSG00000175189 INSR insulin receptor 3643 ENSG00000171105 IRS2 insulin receptor substrate 2 8660 ENSG00000185950 ITGA6 integrin, alpha 6 3655 ENSG00000091409 ITPRIPL2 inositol 1,4,5-trisphosphate receptor interacting 162073 ENSG00000205730 protein-like 2 KARS lysyl-tRNA synthetase 3735 ENSG00000065427 KHDRBS1 KH domain containing, RNA binding, signal transduction 10657 ENSG00000121774 associated 1 KIF5B kinesin family member 5B 3799 ENSG00000170759 LARP1 La ribonucleoprotein domain family, member 1 23367 ENSG00000155506 LDHB lactate dehydrogenase B 3945 ENSG00000111716 LGALS3 lectin, galactoside-binding, soluble, 3 3958 ENSG00000131981 LHFPL2 lipoma HMGIC fusion partner-like 2 10184 ENSG00000145685 LSM4 LSM4 homolog, U6 small nuclear RNA associated (S. 25804 ENSG00000130520 cerevisiae) MAP1LC3B microtubule-associated protein 1 light chain 3 beta 81631 ENSG00000140941 MATR3 matrin 3 9782 ENSG00000015479 MCL1 myeloid cell leukemia sequence 1 (BCL2-related) 4170 ENSG00000143384 MGEA5 meningioma expressed antigen 5 (hyaluronidase) 10724 ENSG00000198408 MLLT4 myeloid/lymphoid or mixed-lineage leukemia (trithorax 4301 ENSG00000130396 homolog, Drosophila); translocated to, 4 MPRIP myosin phosphatase Rho interacting protein 23164 ENSG00000133030 MRC1 mannose receptor, C type 1 4360 ENSG00000120586 MRFAP1L1 Morf4 family associated protein 1-like 1 114932 ENSG00000178988 MTPN myotrophin 136319 ENSG00000105887 MYADM myeloid-associated differentiation marker 91663 ENSG00000179820 NDUFB4 NADH dehydrogenase (ubiquinone) 1 beta subcomplex, 4710 ENSG00000065518 4, 15kDa NHP2L1 NHP2 non-histone chromosome protein 2-like 1 (S. 4809 ENSG00000100138 cerevisiae) NPTN neuroplastin 27020 ENSG00000156642 NUB1 negative regulator of ubiquitin-like proteins 1 51667 ENSG00000013374 NUCKS1 nuclear casein kinase and cyclin-dependent kinase 64710 ENSG00000069275 substrate 1 OXSR1 oxidative-stress responsive 1 9943 ENSG00000172939 PAFAH1B1 platelet-activating factor acetylhydrolase 1b, regulatory 5048 ENSG00000007168 subunit 1 (45kDa) PALLD palladin, cytoskeletal associated protein 23022 ENSG00000129116 PDCD4 programmed cell death 4 (neoplastic transformation 27250 ENSG00000150593 inhibitor)

173 PDGFA platelet-derived growth factor alpha polypeptide 5154 ENSG00000197461 PDIA3 protein disulfide isomerase family A, member 3 2923 ENSG00000167004 PDIA4 protein disulfide isomerase family A, member 4 9601 ENSG00000155660 PFN2 profilin 2 5217 ENSG00000070087 PGAM1 phosphoglycerate mutase 1 (brain) 5223 ENSG00000171314 PHLDB2 pleckstrin homology-like domain, family B, member 2 90102 ENSG00000144824 PICALM phosphatidylinositol binding clathrin assembly protein 8301 ENSG00000073921 PIP4K2C phosphatidylinositol-5-phosphate 4-kinase, type II, 79837 ENSG00000166908 gamma POLR2B polymerase (RNA) II (DNA directed) polypeptide B, 5431 ENSG00000047315 140kDa PPIC peptidylprolyl isomerase C (cyclophilin C) 5480 ENSG00000168938 PSMA2 proteasome (prosome, macropain) subunit, alpha type, 5683 ENSG00000106588,ENSG00000256646 2 PSMD14 proteasome (prosome, macropain) 26S subunit, non- 10213 ENSG00000115233 ATPase, 14 PSMD4 proteasome (prosome, macropain) 26S subunit, non- 5710 ENSG00000159352 ATPase, 4 PSME4 proteasome (prosome, macropain) activator subunit 4 23198 ENSG00000068878 PTGES prostaglandin E synthase 9536 ENSG00000148344 PTGS2 prostaglandin-endoperoxide synthase 2 (prostaglandin 5743 ENSG00000073756 G/H synthase and cyclooxygenase) PURB purine-rich element binding protein B 5814 ENSG00000146676 RAB1A RAB1A, member RAS oncogene family 5861 ENSG00000138069 RALA v-ral simian leukemia viral oncogene homolog A (ras 5898 ENSG00000006451 related) RAN RAN, member RAS oncogene family 5901 ENSG00000132341 RAP2A RAP2A, member of RAS oncogene family 5911 ENSG00000125249 RBM39 RNA binding motif protein 39 9584 ENSG00000131051 RBMS1 RNA binding motif, single stranded interacting protein 1 5937 ENSG00000153250 RCAN1 regulator of calcineurin 1 1827 ENSG00000159200 RDH10 retinol dehydrogenase 10 (all-trans) 157506 ENSG00000121039 RPS21 ribosomal protein S21 6227 ENSG00000171858 RREB1 ras responsive element binding protein 1 6239 ENSG00000124782 RSL24D1 ribosomal L24 domain containing 1 51187 ENSG00000137876 SAR1A SAR1 homolog A (S. cerevisiae) 56681 ENSG00000079332 SEC13 SEC13 homolog (S. cerevisiae) 6396 ENSG00000157020 SEPT2 septin 2 4735 ENSG00000168385 SF3B14 splicing factor 3B, 14 kDa subunit 51639 ENSG00000115128 SFPQ splicing factor proline/glutamine-rich 6421 ENSG00000116560 SKP1 S-phase kinase-associated protein 1 6500 ENSG00000113558 SKP2 S-phase kinase-associated protein 2, E3 ubiquitin 6502 ENSG00000145604 protein ligase SNRPE small nuclear ribonucleoprotein polypeptide E 6635 ENSG00000182004 SNX3 sorting nexin 3 8724 ENSG00000112335 SSB Sjogren syndrome antigen B (autoantigen La) 6741 ENSG00000138385 SSR1 signal sequence receptor, alpha 6745 ENSG00000124783 SSR2 signal sequence receptor, beta (translocon-associated 6746 ENSG00000163479 protein beta)

174 STAT2 signal transducer and activator of transcription 2, 6773 ENSG00000170581 113kDa STC1 stanniocalcin 1 6781 ENSG00000159167 SYNCRIP synaptotagmin binding, cytoplasmic RNA interacting 10492 ENSG00000135316 protein TALDO1 transaldolase 1 6888 ENSG00000177156 TAPBP TAP binding protein (tapasin) 6892 ENSG00000231925 TARS threonyl-tRNA synthetase 6897 ENSG00000113407 TAX1BP3 Tax1 (human T-cell leukemia virus type I) binding 30851 ENSG00000213977 protein 3 TFAP2A transcription factor AP-2 alpha (activating enhancer 7020 ENSG00000137203 binding protein 2 alpha) TJP1 tight junction protein 1 7082 ENSG00000104067 TM2D3 TM2 domain containing 3 80213 ENSG00000184277 TPBG trophoblast glycoprotein 7162 ENSG00000146242 TPM2 tropomyosin 2 (beta) 7169 ENSG00000198467 TPM4 tropomyosin 4 7171 ENSG00000167460 TRIM8 tripartite motif containing 8 81603 ENSG00000171206 TSPAN13 tetraspanin 13 27075 ENSG00000106537 TUBB6 tubulin, beta 6 class V 84617 ENSG00000176014 TXNL1 thioredoxin-like 1 9352 ENSG00000091164 U2AF1 U2 small nuclear RNA auxiliary factor 1 7307 ENSG00000160201 UBE2D3 ubiquitin-conjugating enzyme E2D 3 7323 ENSG00000109332 VCAN versican 1462 ENSG00000038427 YTHDC1 YTH domain containing 1 91746 ENSG00000083896 YTHDF2 YTH domain family, member 2 51441 ENSG00000198492 ZBTB4 zinc finger and BTB domain containing 4 57659 ENSG00000174282 ZFAND6 zinc finger, AN1-type domain 6 54469 ENSG00000086666 ZFP36L1 zinc finger protein 36, C3H type-like 1 677 ENSG00000185650

HSPE = Histamine Specific genes in PE (elevated histamine regulated significant genes also expressed consistently in PE placentae); PE = Pre-eclampsia

175 Appendix 7: Down-Regulated HSPE genes Gene Symbol Gene Name EntrezGene Ensembl ABCB5 ATP-binding cassette, sub-family B (MDR/TAP), member 5 340273 ENSG00000004846 ABRA actin-binding Rho activating protein 137735 ENSG00000174429 ACSM2B acyl-CoA synthetase medium-chain family member 2B 348158 ENSG00000066813 ADCYAP1R1 adenylate cyclase activating polypeptide 1 (pituitary) 117 ENSG00000078549 receptor type I APOF apolipoprotein F 319 ENSG00000175336 ARG1 arginase, liver 383 ENSG00000118520 ASB12 ankyrin repeat and SOCS box containing 12 142689 ENSG00000198881 ASB5 ankyrin repeat and SOCS box containing 5 140458 ENSG00000164122 B3GALT2 UDP-Gal:betaGlcNAc beta 1,3-galactosyltransferase, 8707 ENSG00000162630 polypeptide 2 BCL2L14 BCL2-like 14 (apoptosis facilitator) 79370 ENSG00000121380 C12orf40 chromosome 12 open reading frame 40 283461 ENSG00000180116 C14orf39 chromosome 14 open reading frame 39 317761 ENSG00000179008 C15orf43 chromosome 15 open reading frame 43 145645 ENSG00000167014 C1orf110 chromosome 1 open reading frame 110 339512 ENSG00000185860 C3orf15 chromosome 3 open reading frame 15 89876 ENSG00000183833 C6orf10 chromosome 6 open reading frame 10 10665 ENSG00000204296 C8orf34 chromosome 8 open reading frame 34 116328 ENSG00000165084 CA6 carbonic anhydrase VI 765 ENSG00000131686 CADPS Ca++-dependent secretion activator 8618 ENSG00000163618 CCDC67 coiled-coil domain containing 67 159989 ENSG00000165325 CFTR cystic fibrosis transmembrane conductance regulator 1080 ENSG00000001626 (ATP-binding cassette sub-family C, member 7) CNBD1 cyclic nucleotide binding domain containing 1 168975 ENSG00000176571 CNTNAP4 contactin associated protein-like 4 85445 ENSG00000152910 CNTNAP5 contactin associated protein-like 5 129684 ENSG00000155052 CPEB2 cytoplasmic polyadenylation element binding protein 2 132864 ENSG00000137449 CRB1 crumbs homolog 1 (Drosophila) 23418 ENSG00000134376 CRYGB crystallin, gamma B 1419 ENSG00000182187 CSN2 casein beta 1447 ENSG00000135222 DEFB126 defensin, beta 126 81623 NA EFCAB6 EF-hand calcium binding domain 6 64800 ENSG00000186976 EMCN endomucin 51705 ENSG00000164035 ENAM enamelin 10117 ENSG00000132464 ESRRB estrogen-related receptor beta 2103 ENSG00000119715 FABP6 fatty acid binding protein 6, ileal 2172 ENSG00000170231 FCRL5 Fc receptor-like 5 83416 ENSG00000143297 FLG filaggrin 2312 ENSG00000143631 FOXA1 forkhead box A1 3169 ENSG00000129514 FSTL5 follistatin-like 5 56884 ENSG00000168843 GC group-specific component (vitamin D binding protein) 2638 ENSG00000145321 GCM2 glial cells missing homolog 2 (Drosophila) 9247 ENSG00000124827 GLRA3 glycine receptor, alpha 3 8001 ENSG00000145451 GLYAT glycine-N-acyltransferase 10249 ENSG00000149124 GPM6A glycoprotein M6A 2823 ENSG00000150625

176 GPR17 G protein-coupled receptor 17 2840 ENSG00000144230 GPR22 G protein-coupled receptor 22 2845 ENSG00000172209 GRIN2B glutamate receptor, ionotropic, N-methyl D-aspartate 2B 2904 ENSG00000150086 HAL histidine ammonia-lyase 3034 ENSG00000084110 HAO1 hydroxyacid oxidase (glycolate oxidase) 1 54363 ENSG00000101323 HEPACAM2 HEPACAM family member 2 253012 ENSG00000188175 HHLA2 HERV-H LTR-associating 2 11148 ENSG00000114455 IGFBP1 insulin-like growth factor binding protein 1 3484 ENSG00000146678 INTS4 integrator complex subunit 4 92105 ENSG00000149262 IQGAP2 IQ motif containing GTPase activating protein 2 10788 ENSG00000145703 KCNK10 potassium channel, subfamily K, member 10 54207 ENSG00000100433 KLF12 Kruppel-like factor 12 11278 ENSG00000118922 KRT38 keratin 38 8687 ENSG00000171360 LAMA2 laminin, alpha 2 3908 ENSG00000196569 LHX9 LIM homeobox 9 56956 ENSG00000143355 LMO3 LIM domain only 3 (rhombotin-like 2) 55885 ENSG00000048540 LOC400891 chromosome 14 open reading frame 166B pseudogene 400891 NA MAGEC2 melanoma antigen family C, 2 51438 ENSG00000046774 MAGEE2 melanoma antigen family E, 2 139599 ENSG00000186675 MC2R melanocortin 2 receptor (adrenocorticotropic hormone) 4158 ENSG00000185231 MIPOL1 mirror-image polydactyly 1 145282 ENSG00000151338 MPL myeloproliferative leukemia virus oncogene 4352 ENSG00000117400 MTMR8 myotubularin related protein 8 55613 ENSG00000102043 MYO3B myosin IIIB 140469 ENSG00000071909 MYT1L myelin transcription factor 1-like 23040 ENSG00000186487 NCAM1 neural cell adhesion molecule 1 4684 ENSG00000149294 NCR2 natural cytotoxicity triggering receptor 2 9436 ENSG00000096264 NDST4 N-deacetylase/N-sulfotransferase (heparan glucosaminyl) 64579 ENSG00000138653 4 NOVA1 neuro-oncological ventral antigen 1 4857 ENSG00000139910 NPBWR2 neuropeptides B/W receptor 2 2832 ENSG00000125522 OLFM3 olfactomedin 3 118427 ENSG00000118733 OPRM1 opioid receptor, mu 1 4988 ENSG00000112038 OR2J2 olfactory receptor, family 2, subfamily J, member 2 26707 ENSG00000204700 OR5V1 olfactory receptor, family 5, subfamily V, member 1 81696 ENSG00000243729 P2RY10 purinergic receptor P2Y, G-protein coupled, 10 27334 ENSG00000078589 PCDH20 protocadherin 20 64881 ENSG00000197991 PDE11A phosphodiesterase 11A 50940 ENSG00000128655 PER3 period homolog 3 (Drosophila) 8863 ENSG00000049246 PMS1 PMS1 postmeiotic segregation increased 1 (S. cerevisiae) 5378 ENSG00000064933 POF1B premature ovarian failure, 1B 79983 ENSG00000124429 POU1F1 POU class 1 homeobox 1 5449 ENSG00000064835 PPP1R3A protein phosphatase 1, regulatory subunit 3A 5506 ENSG00000154415 PPP3R2 protein phosphatase 3, regulatory subunit B, beta 5535 ENSG00000188386 PREX2 phosphatidylinositol-3,4,5-trisphosphate-dependent Rac 80243 ENSG00000046889 exchange factor 2 PTCHD1 patched domain containing 1 139411 ENSG00000165186 PTH parathyroid hormone 5741 ENSG00000152266

177 PTN pleiotrophin 5764 ENSG00000105894 RALYL RALY RNA binding protein-like 138046 ENSG00000184672 RNF165 ring finger protein 165 494470 ENSG00000141622 SCEL sciellin 8796 ENSG00000136155 SCML4 sex comb on midleg-like 4 (Drosophila) 256380 ENSG00000146285 SCN2B sodium channel, voltage-gated, type II, beta subunit 6327 ENSG00000149575 SDR16C5 short chain dehydrogenase/reductase family 16C, 195814 ENSG00000170786 member 5 SERPINB4 serpin peptidase inhibitor, clade B (ovalbumin), member 4 6318 ENSG00000206073 SFRP4 secreted frizzled-related protein 4 6424 ENSG00000106483 SIM1 single-minded homolog 1 (Drosophila) 6492 ENSG00000112246 SLC13A1 solute carrier family 13 (sodium/sulfate symporters), 6561 ENSG00000081800 member 1 SLC17A1 solute carrier family 17 (sodium phosphate), member 1 6568 ENSG00000124568 SLC17A8 solute carrier family 17 (sodium-dependent inorganic 246213 ENSG00000179520 phosphate cotransporter), member 8 SLC26A7 solute carrier family 26, member 7 115111 ENSG00000147606 SLC5A7 solute carrier family 5 (choline transporter), member 7 60482 ENSG00000115665 SOX6 SRY (sex determining region Y)-box 6 55553 ENSG00000110693 SRP72 signal recognition particle 72kDa 6731 ENSG00000174780 TAS2R16 taste receptor, type 2, member 16 50833 ENSG00000128519 TCTE3 t-complex-associated-testis-expressed 3 6991 ENSG00000184786 TMED5 transmembrane emp24 protein transport domain 50999 ENSG00000117500 containing 5 TMEM67 transmembrane protein 67 91147 ENSG00000164953 TPPP2 tubulin polymerization-promoting protein family member 122664 ENSG00000179636 2 TRDN triadin 10345 ENSG00000186439 TRPC5 transient receptor potential cation channel, subfamily C, 7224 ENSG00000072315 member 5 TRPM1 transient receptor potential cation channel, subfamily M, 4308 ENSG00000134160 member 1 TSHR thyroid stimulating hormone receptor 7253 ENSG00000165409 UGT2A3 UDP glucuronosyltransferase 2 family, polypeptide A3 79799 ENSG00000135220 UNC5C unc-5 homolog C (C. elegans) 8633 ENSG00000182168 UPP2 uridine phosphorylase 2 151531 ENSG00000007001 VASH2 vasohibin 2 79805 ENSG00000143494 ZDHHC15 zinc finger, DHHC-type containing 15 158866 ENSG00000102383 ZNF407 zinc finger protein 407 55628 ENSG00000215421 ZNF780B zinc finger protein 780B 163131 ENSG00000128000 HSPE = Histamine Specific genes in PE (elevated histamine regulated significant genes also expressed consistently in PE placentae); PE = Pre-eclampsia

178 Appendix 8: Validation Genes Primer sequences for RT-qPCR Target Primer sequence (5’ to 3’) Product GenBank ® length (bp) accession number L19 F: GCGGAAGGGTACAGCCAAT 140 NM_000981.3 R: GCAGCCGGCGCAAA AKR1C1 F: ATTTGCCAGCCAGGCTAGTG 122 NM_001353.5 R: ACTTTTAGGACCTCTGCAGGC EREG F: CTGCCTGGGTTTCCATCTTCT 123 NM_001412.2 R: GCCACACGTGGATTGTCTTC IGFBP5 F: AAAGCAGTGCAAACCTTCCC 138 NM_001353.5 R: CACTCAACGTTGCTGCTGTC

AKR1C1, aldo-ketose reductase family 1 member C1; EREG, epiregulin; IGFBP5, insulin-like growth factor binding protein 5; RT-qPCR, real-time quantitative polymerase chain reaction;

Appendix 9: RT-qPCR Validation results relative to L19

179

180

181